Hackaton.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Import of libraries and data sets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from feature_extraction import *\n",
    "from nice_plots import *\n",
    "from imblearn.over_sampling import SMOTE\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.model_selection import RandomizedSearchCV,StratifiedKFold\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.decomposition import PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import data set\n",
    "df = pd.read_csv(\"train.csv\")\n",
    "test_df =  pd.read_csv(\"test.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Our setup : we have for 663 observations, in different potions and different observation times, 24 devices recording measurements in the form of timeseries of length 2048."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>main.pos</th>\n",
       "      <th>fun.pos</th>\n",
       "      <th>when</th>\n",
       "      <th>y</th>\n",
       "      <th>E1A1</th>\n",
       "      <th>E1A2</th>\n",
       "      <th>E1A3</th>\n",
       "      <th>E1A4</th>\n",
       "      <th>E1A5</th>\n",
       "      <th>...</th>\n",
       "      <th>M2V2039</th>\n",
       "      <th>M2V2040</th>\n",
       "      <th>M2V2041</th>\n",
       "      <th>M2V2042</th>\n",
       "      <th>M2V2043</th>\n",
       "      <th>M2V2044</th>\n",
       "      <th>M2V2045</th>\n",
       "      <th>M2V2046</th>\n",
       "      <th>M2V2047</th>\n",
       "      <th>M2V2048</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>LHCB-P18</td>\n",
       "      <td>QSCB-18-C2</td>\n",
       "      <td>2017-06</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.576589</td>\n",
       "      <td>-0.951065</td>\n",
       "      <td>0.944338</td>\n",
       "      <td>-0.086406</td>\n",
       "      <td>0.253089</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.017011</td>\n",
       "      <td>-0.101842</td>\n",
       "      <td>0.027109</td>\n",
       "      <td>0.168339</td>\n",
       "      <td>0.152390</td>\n",
       "      <td>0.000410</td>\n",
       "      <td>-0.126250</td>\n",
       "      <td>0.072403</td>\n",
       "      <td>0.214993</td>\n",
       "      <td>0.184996</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>LHCB-6</td>\n",
       "      <td>QSCB-6-CP6</td>\n",
       "      <td>2018-09</td>\n",
       "      <td>1</td>\n",
       "      <td>0.970694</td>\n",
       "      <td>1.524991</td>\n",
       "      <td>1.813341</td>\n",
       "      <td>0.447012</td>\n",
       "      <td>-1.038631</td>\n",
       "      <td>...</td>\n",
       "      <td>0.106717</td>\n",
       "      <td>-0.057328</td>\n",
       "      <td>-0.146602</td>\n",
       "      <td>0.069251</td>\n",
       "      <td>0.351492</td>\n",
       "      <td>0.106101</td>\n",
       "      <td>0.192691</td>\n",
       "      <td>0.167834</td>\n",
       "      <td>-0.170475</td>\n",
       "      <td>-0.236751</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>LHCB-8</td>\n",
       "      <td>QSCB-8-CP7</td>\n",
       "      <td>2022-11</td>\n",
       "      <td>1</td>\n",
       "      <td>1.065262</td>\n",
       "      <td>0.461292</td>\n",
       "      <td>0.416869</td>\n",
       "      <td>0.460564</td>\n",
       "      <td>-0.019386</td>\n",
       "      <td>...</td>\n",
       "      <td>0.198017</td>\n",
       "      <td>-0.044788</td>\n",
       "      <td>0.011786</td>\n",
       "      <td>-0.037881</td>\n",
       "      <td>-0.152690</td>\n",
       "      <td>-0.104267</td>\n",
       "      <td>0.037444</td>\n",
       "      <td>0.489454</td>\n",
       "      <td>0.268134</td>\n",
       "      <td>-0.117367</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>LHCB-8</td>\n",
       "      <td>QSCB-8-CP1</td>\n",
       "      <td>2023-04</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.341942</td>\n",
       "      <td>0.614190</td>\n",
       "      <td>0.904430</td>\n",
       "      <td>0.727268</td>\n",
       "      <td>0.187494</td>\n",
       "      <td>...</td>\n",
       "      <td>0.168140</td>\n",
       "      <td>-0.071131</td>\n",
       "      <td>0.067982</td>\n",
       "      <td>-0.023509</td>\n",
       "      <td>-0.028273</td>\n",
       "      <td>0.000943</td>\n",
       "      <td>-0.018128</td>\n",
       "      <td>0.013366</td>\n",
       "      <td>0.015902</td>\n",
       "      <td>0.043247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>LHCB-P18</td>\n",
       "      <td>QSCB-18-C1</td>\n",
       "      <td>2017-04</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.569563</td>\n",
       "      <td>0.368497</td>\n",
       "      <td>-0.269832</td>\n",
       "      <td>-0.179988</td>\n",
       "      <td>-0.065328</td>\n",
       "      <td>...</td>\n",
       "      <td>0.082521</td>\n",
       "      <td>0.105978</td>\n",
       "      <td>0.020868</td>\n",
       "      <td>0.173724</td>\n",
       "      <td>-0.037431</td>\n",
       "      <td>0.147695</td>\n",
       "      <td>0.038233</td>\n",
       "      <td>-0.047549</td>\n",
       "      <td>0.101227</td>\n",
       "      <td>-0.109760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>658</th>\n",
       "      <td>1012</td>\n",
       "      <td>LHCB-4</td>\n",
       "      <td>QSCB-4-CP1</td>\n",
       "      <td>2019-03</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.094653</td>\n",
       "      <td>-0.273495</td>\n",
       "      <td>0.927676</td>\n",
       "      <td>-0.244089</td>\n",
       "      <td>0.394102</td>\n",
       "      <td>...</td>\n",
       "      <td>0.076421</td>\n",
       "      <td>-0.017458</td>\n",
       "      <td>0.113222</td>\n",
       "      <td>0.081531</td>\n",
       "      <td>-0.118022</td>\n",
       "      <td>-0.068159</td>\n",
       "      <td>-0.160683</td>\n",
       "      <td>-0.054098</td>\n",
       "      <td>-0.043842</td>\n",
       "      <td>0.079474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>659</th>\n",
       "      <td>1013</td>\n",
       "      <td>LHCB-6</td>\n",
       "      <td>QSCB-6-CP1</td>\n",
       "      <td>2022-05</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.447106</td>\n",
       "      <td>-0.074231</td>\n",
       "      <td>0.331543</td>\n",
       "      <td>0.150676</td>\n",
       "      <td>-0.050445</td>\n",
       "      <td>...</td>\n",
       "      <td>0.137185</td>\n",
       "      <td>0.240015</td>\n",
       "      <td>0.169615</td>\n",
       "      <td>0.068595</td>\n",
       "      <td>0.097196</td>\n",
       "      <td>-0.020113</td>\n",
       "      <td>-0.010522</td>\n",
       "      <td>0.063634</td>\n",
       "      <td>0.096408</td>\n",
       "      <td>-0.085637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>660</th>\n",
       "      <td>1014</td>\n",
       "      <td>LHCB-P18</td>\n",
       "      <td>QSCB-18-C2</td>\n",
       "      <td>2018-04</td>\n",
       "      <td>1</td>\n",
       "      <td>0.264532</td>\n",
       "      <td>1.554129</td>\n",
       "      <td>-1.447155</td>\n",
       "      <td>0.442756</td>\n",
       "      <td>2.655931</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.037920</td>\n",
       "      <td>0.087096</td>\n",
       "      <td>-0.399536</td>\n",
       "      <td>-0.121802</td>\n",
       "      <td>-0.027615</td>\n",
       "      <td>-0.190010</td>\n",
       "      <td>-0.099357</td>\n",
       "      <td>-0.026138</td>\n",
       "      <td>-0.008852</td>\n",
       "      <td>-0.108788</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>661</th>\n",
       "      <td>1015</td>\n",
       "      <td>LHCB-4</td>\n",
       "      <td>QSCB-4-CP7</td>\n",
       "      <td>2017-01</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.522473</td>\n",
       "      <td>-0.007026</td>\n",
       "      <td>0.361770</td>\n",
       "      <td>-0.630256</td>\n",
       "      <td>-0.851504</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.256542</td>\n",
       "      <td>-0.215720</td>\n",
       "      <td>-0.006428</td>\n",
       "      <td>0.037543</td>\n",
       "      <td>0.120194</td>\n",
       "      <td>0.155707</td>\n",
       "      <td>0.156359</td>\n",
       "      <td>0.280801</td>\n",
       "      <td>0.155129</td>\n",
       "      <td>0.176686</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>662</th>\n",
       "      <td>1017</td>\n",
       "      <td>LHCB-P18</td>\n",
       "      <td>QSCB-18-C2</td>\n",
       "      <td>2017-04</td>\n",
       "      <td>1</td>\n",
       "      <td>1.817682</td>\n",
       "      <td>0.419835</td>\n",
       "      <td>0.547951</td>\n",
       "      <td>-1.162066</td>\n",
       "      <td>-1.240259</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.018203</td>\n",
       "      <td>0.151254</td>\n",
       "      <td>0.007341</td>\n",
       "      <td>0.115778</td>\n",
       "      <td>0.089899</td>\n",
       "      <td>0.123921</td>\n",
       "      <td>-0.033705</td>\n",
       "      <td>0.111735</td>\n",
       "      <td>-0.057554</td>\n",
       "      <td>0.137242</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>663 rows × 49157 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       id  main.pos     fun.pos     when  y      E1A1      E1A2      E1A3  \\\n",
       "0       1  LHCB-P18  QSCB-18-C2  2017-06  1 -0.576589 -0.951065  0.944338   \n",
       "1       2    LHCB-6  QSCB-6-CP6  2018-09  1  0.970694  1.524991  1.813341   \n",
       "2       3    LHCB-8  QSCB-8-CP7  2022-11  1  1.065262  0.461292  0.416869   \n",
       "3       4    LHCB-8  QSCB-8-CP1  2023-04  1 -0.341942  0.614190  0.904430   \n",
       "4       5  LHCB-P18  QSCB-18-C1  2017-04  1 -0.569563  0.368497 -0.269832   \n",
       "..    ...       ...         ...      ... ..       ...       ...       ...   \n",
       "658  1012    LHCB-4  QSCB-4-CP1  2019-03  1 -0.094653 -0.273495  0.927676   \n",
       "659  1013    LHCB-6  QSCB-6-CP1  2022-05  1 -0.447106 -0.074231  0.331543   \n",
       "660  1014  LHCB-P18  QSCB-18-C2  2018-04  1  0.264532  1.554129 -1.447155   \n",
       "661  1015    LHCB-4  QSCB-4-CP7  2017-01  1 -0.522473 -0.007026  0.361770   \n",
       "662  1017  LHCB-P18  QSCB-18-C2  2017-04  1  1.817682  0.419835  0.547951   \n",
       "\n",
       "         E1A4      E1A5  ...   M2V2039   M2V2040   M2V2041   M2V2042  \\\n",
       "0   -0.086406  0.253089  ... -0.017011 -0.101842  0.027109  0.168339   \n",
       "1    0.447012 -1.038631  ...  0.106717 -0.057328 -0.146602  0.069251   \n",
       "2    0.460564 -0.019386  ...  0.198017 -0.044788  0.011786 -0.037881   \n",
       "3    0.727268  0.187494  ...  0.168140 -0.071131  0.067982 -0.023509   \n",
       "4   -0.179988 -0.065328  ...  0.082521  0.105978  0.020868  0.173724   \n",
       "..        ...       ...  ...       ...       ...       ...       ...   \n",
       "658 -0.244089  0.394102  ...  0.076421 -0.017458  0.113222  0.081531   \n",
       "659  0.150676 -0.050445  ...  0.137185  0.240015  0.169615  0.068595   \n",
       "660  0.442756  2.655931  ... -0.037920  0.087096 -0.399536 -0.121802   \n",
       "661 -0.630256 -0.851504  ... -0.256542 -0.215720 -0.006428  0.037543   \n",
       "662 -1.162066 -1.240259  ... -0.018203  0.151254  0.007341  0.115778   \n",
       "\n",
       "      M2V2043   M2V2044   M2V2045   M2V2046   M2V2047   M2V2048  \n",
       "0    0.152390  0.000410 -0.126250  0.072403  0.214993  0.184996  \n",
       "1    0.351492  0.106101  0.192691  0.167834 -0.170475 -0.236751  \n",
       "2   -0.152690 -0.104267  0.037444  0.489454  0.268134 -0.117367  \n",
       "3   -0.028273  0.000943 -0.018128  0.013366  0.015902  0.043247  \n",
       "4   -0.037431  0.147695  0.038233 -0.047549  0.101227 -0.109760  \n",
       "..        ...       ...       ...       ...       ...       ...  \n",
       "658 -0.118022 -0.068159 -0.160683 -0.054098 -0.043842  0.079474  \n",
       "659  0.097196 -0.020113 -0.010522  0.063634  0.096408 -0.085637  \n",
       "660 -0.027615 -0.190010 -0.099357 -0.026138 -0.008852 -0.108788  \n",
       "661  0.120194  0.155707  0.156359  0.280801  0.155129  0.176686  \n",
       "662  0.089899  0.123921 -0.033705  0.111735 -0.057554  0.137242  \n",
       "\n",
       "[663 rows x 49157 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# have a look\n",
    "df"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Functions for feature extraction"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Max power spectrum (FFT)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The **Fast Fourier Transform (FFT)** is an efficient algorithm used to compute the Discrete Fourier Transform (DFT) of a signal. The DFT is a mathematical transformation that converts a time-domain signal into its frequency-domain representation. In simpler terms, the FFT converts a signal from the time domain to the frequency domain, providing information about the various frequencies present in the signal.\n",
    "\n",
    "Mathematically, for a time series $\\{x[n]\\}$ of length N, the DFT is defined as:\n",
    "\n",
    "$X[k] = \\sum {x[n] * exp(-j * 2 * \\pi * k * n / N)}$\n",
    "\n",
    "where $X[k]$ is the complex-valued representation of the signal at frequency bin k (ranging from 0 to N-1), and j is the imaginary unit.\n",
    "\n",
    "The FFT algorithm significantly speeds up the computation of the DFT by exploiting symmetries and reducing the number of computations required, making it practical for real-world applications.\n",
    "\n",
    "Now, let's explore why the Fast Fourier Transform can be useful in anomaly detection:\n",
    "\n",
    "1. Frequency Analysis: The FFT provides a frequency representation of the signal, highlighting the presence and magnitude of different frequency components. Anomalies might be characterized by unusual or unexpected frequencies in the signal.\n",
    "\n",
    "2. Spectral Leakage: When analyzing periodic signals or events with a specific frequency, the FFT can show clear spikes or peaks at the corresponding frequency bins. Spectral leakage (a phenomenon in FFT) might cause the energy from a frequency to spread to neighboring bins. Detecting such energy spread can help in identifying potential anomalies.\n",
    "\n",
    "3. Detecting Unexpected Frequencies: Anomalies may introduce new frequencies not present in the normal behavior. By observing significant energy spikes at unusual frequencies, you can detect deviations from the expected patterns.\n",
    "\n",
    "4. Frequency Trends: FFT can help detect trends in the frequency domain, such as a gradual increase or decrease in certain frequency components, which may indicate a change in the system's behavior.\n",
    "\n",
    "5. Noise Analysis: In some cases, anomalies can manifest as noise or interference in the signal. FFT can help distinguish between the noise and meaningful frequency components, making it easier to detect anomalies.\n",
    "\n",
    "6. Feature Extraction: In many anomaly detection algorithms, the FFT is used as a preprocessing step to extract frequency-domain features. These features can be fed into machine learning models or anomaly detection algorithms for classification and anomaly identification.\n",
    "\n",
    "7. Comparison with Normal Behavior: By analyzing the FFT of the signal during normal operation, you can establish a baseline or reference for the expected frequency distribution. During anomaly detection, any deviation from this baseline can be considered an anomaly."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sample_ts= df.iloc[0,5:2053] # take a sample time series\n",
    "plot_magnitudes(sample_ts)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Thus our first approach involves switching from time space to frequency space using the Fast Fourier Transform.\n",
    "\n",
    "In our case we therefore choose, despite the fact that the temporal extent of the individual timeseries is **unknown**, to evaluate the power spectrum and for the first half of the frequencies and extract features in the following way:\n",
    "1. calculate the power spectrum by means of the **fourier transform**;\n",
    "2. take only the real part of the power spectrum (applying abs);\n",
    "3. do the square of the magnitudes;\n",
    "4. divide the frequency axis (x-axis) into bins;\n",
    "5. for each bin take the maximum power spectrum (magnitude)."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This will be the output of the function we are going to use for the extraction of this features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_peaks(sample_ts)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Power spectral density"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Power spectral density (PSD) shows how the power of a signal is distributed over frequencies. Periodograms are a particular version of PSDs applied for discrete-time signals and we are going to use them to evaluate our PSD. We decide to extract 3 main summaries form this distribution since seems to be important for tasks related to anomaly detection and vibration/accelerometer analysis:\n",
    "1. median of the PSD;\n",
    "2. mean absolute deviation (generally nice in combo with the median);\n",
    "3. skewness of the PSD."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjsAAAHHCAYAAABZbpmkAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAABmrUlEQVR4nO3deXwTZf4H8E+SpundUqAthXIroCLnUlFOuXE5FGERRBAX1B/oT1gvdlUorKIsCh4cKz8BDxDUFVwVgXJfBbnvq5UCFVqO0pa2NM3x/P6IDZk2PVKSyWTyeb9eaDOZzDzzzeTJN8/zzDMaIYQAERERkUppvV0AIiIiIk9iskNERESqxmSHiIiIVI3JDhEREakakx0iIiJSNSY7REREpGpMdoiIiEjVmOwQERGRqjHZISIiIlVjskNEbtGwYUOMGTOmWq9NT0+HRqPB0qVL3Vqm6tBoNJg2bZq3iyHx5Zdfonnz5tDr9YiKipJ9/2PGjEHDhg1l36+c/OEY/RmTHZLN0qVLodFo7P+CgoJw9913Y+LEicjKyvJ28e6Y1WrFF198gcTERERHRyM8PBx33303nnrqKezevdvbxcOuXbswbdo05OTkeLsoPsXbcTt16hTGjBmDJk2aYNGiRfj00089sp9Lly5h2rRpOHTokEe2T+RNAd4uAPmf6dOno1GjRigqKsKOHTuwYMECrFmzBseOHUNISIi3i1dtL774IubNm4dBgwZh5MiRCAgIwOnTp/HLL7+gcePGeOCBB7xavl27diEpKQljxozxSuuAr7h16xYCAm5Xjd6O25YtW2C1WvHhhx+iadOmHtvPpUuXkJSUhIYNG6J169aS5xYtWgSr1eqxfSuBPxyjP2OyQ7Lr168f2rdvDwD461//ipo1a+KDDz7ADz/8gCeeeMLLpSuf1WpFcXExgoKCyjyXlZWF+fPnY9y4cWV+ec+dOxdXr16Vq5huUdGxqp3SjvnKlSsA4NUEVa/Xe23fnlZQUIDQ0FBVHyOxG4sU4OGHHwYAnDt3DgBgNpsxY8YMNGnSBAaDAQ0bNsTf//53GI1G+2smT56MmjVrQghhX/bCCy9Ao9Hgo48+si/LysqCRqPBggUL7MuMRiOmTp2Kpk2bwmAwICEhAa+++qpk+4Bt7MbEiROxbNky3HvvvTAYDFi7dq3TYzh37hyEEHjooYfKPKfRaBATE2N/XNKdt23bNjz77LOoWbMmIiIi8NRTT+HGjRtlXv/LL7+gc+fOCA0NRXh4OB555BEcP368zHqnTp3CsGHDULt2bQQHB6NZs2b4xz/+AQCYNm0aXnnlFQBAo0aN7F2J6enplR7r7Nmz8eCDD6JmzZoIDg5Gu3bt8N133zmNQ1Xk5ORgzJgxiIyMRFRUFEaPHl1uF9GpU6fw+OOPIzo6GkFBQWjfvj3++9//StYpiefOnTsxefJk1K5dG6GhoXj00UfLJJn79u1Dnz59UKtWLQQHB6NRo0YYO3asZB3HMTsVxa1r165o1aqV03I3a9YMffr0qTQW8+fPt8c7Pj4eEyZMkMSiYcOGmDp1KgCgdu3alY4nOnLkCMaMGYPGjRsjKCgIcXFxGDt2LK5fv15hObZs2YI//elPAICnn37afpwlY6hKj2cpGWM1e/ZszJs3D40bN0ZISAh69+6NixcvQgiBGTNmoF69eggODsagQYOQnZ1dZr9VObczMzPx9NNPo169ejAYDKhTpw4GDRpkP3dd2daYMWMQFhaGtLQ09O/fH+Hh4Rg5cqTTYwRsSf/cuXNx7733IigoCLGxsXj22WfLfE6rcl6Rd7Flh7wuLS0NAFCzZk0Attaezz//HI8//jj+9re/Yc+ePZg5cyZOnjyJVatWAQA6d+6MOXPm4Pjx47jvvvsAANu3b4dWq8X27dvx4osv2pcBQJcuXQDYKq+BAwdix44dGD9+PFq0aIGjR49izpw5OHPmDFavXi0p26ZNm/DNN99g4sSJqFWrVrkDGBs0aAAA+PbbbzF06NAqdcdNnDgRUVFRmDZtGk6fPo0FCxbg/Pnz2LJlCzQaDQDbwNTRo0ejT58+eO+991BYWIgFCxagU6dOOHjwoL08R44cQefOnaHX6zF+/Hg0bNgQaWlp+PHHH/H222/jsccew5kzZ/D1119jzpw5qFWrFgDbF2hlx/rhhx9i4MCBGDlyJIqLi7FixQoMHToUP/30Ex555JFKj9OREAKDBg3Cjh078Nxzz6FFixZYtWoVRo8eXWbd48eP46GHHkLdunXx+uuvIzQ0FN988w0GDx6M//znP3j00Ucl67/wwguoUaMGpk6divT0dMydOxcTJ07EypUrAdhaSHr37o3atWvj9ddfR1RUFNLT0/H999+XW96K4jZq1CiMGzcOx44ds5+DALB3716cOXMGb7zxRoWxmDZtGpKSktCzZ088//zz9nNg79692LlzJ/R6PebOnYsvvvgCq1atwoIFCxAWFob777+/3G0mJyfjt99+w9NPP424uDgcP34cn376KY4fP47du3fbz6vSWrRogenTp+Ott97C+PHj0blzZwDAgw8+WOExLFu2DMXFxXjhhReQnZ2NWbNmYdiwYXj44YexZcsWvPbaa0hNTcXHH3+Ml19+GYsXL7a/tqrn9pAhQ3D8+HG88MILaNiwIa5cuYLk5GRcuHDBvk5VtwXYfkz16dMHnTp1wuzZsyv8rD777LNYunQpnn76abz44os4d+4cPvnkExw8eND+HlXnvCIvEEQyWbJkiQAgNmzYIK5evSouXrwoVqxYIWrWrCmCg4NFRkaGOHTokAAg/vrXv0pe+/LLLwsAYtOmTUIIIa5cuSIAiPnz5wshhMjJyRFarVYMHTpUxMbG2l/34osviujoaGG1WoUQQnz55ZdCq9WK7du3S7a/cOFCAUDs3LnTvgyA0Gq14vjx41U6vqeeekoAEDVq1BCPPvqomD17tjh58mS5cWjXrp0oLi62L581a5YAIH744QchhBA3b94UUVFRYty4cZLXZ2ZmisjISMnyLl26iPDwcHH+/HnJuiXHLYQQ//rXvwQAce7cuTJlquhYCwsLJY+Li4vFfffdJx5++GHJ8gYNGojRo0eXeb2j1atXCwBi1qxZ9mVms1l07txZABBLliyxL+/Ro4do2bKlKCoqkhzPgw8+KO666y77spJ49uzZU3K8kyZNEjqdTuTk5AghhFi1apUAIPbu3VthGQGIqVOn2h+XF7ecnBwRFBQkXnvtNcnyF198UYSGhor8/Pxy93HlyhURGBgoevfuLSwWi335J598IgCIxYsX25dNnTpVABBXr16tsNxClH2vhBDi66+/FgDEtm3bKnzt3r17y7wHJUaPHi0aNGhgf3zu3DkBQNSuXdseXyGEmDJligAgWrVqJUwmk335E088IQIDA+3vZVXP7Rs3bggA4l//+le55XblczJ69GgBQLz++uuVHuP27dsFALFs2TLJemvXrpUsr+p5Rd7FbiySXc+ePVG7dm0kJCRg+PDhCAsLw6pVq1C3bl2sWbMGgK2bytHf/vY3AMDPP/8MwPbLunnz5ti2bRsAYOfOndDpdHjllVeQlZWFs2fPArC17HTq1Mn+i/bbb79FixYt0Lx5c1y7ds3+r6QrbfPmzZL9du3aFffcc0+VjmvJkiX45JNP0KhRI6xatQovv/wyWrRogR49euD3338vs/748eMl4wSef/55BAQE2GOQnJyMnJwcPPHEE5Ky6nQ6JCYm2st69epVbNu2DWPHjkX9+vUl+yjvl7wz5R1rcHCw/e8bN24gNzcXnTt3xoEDB6q87RJr1qxBQEAAnn/+efsynU6HF154QbJednY2Nm3ahGHDhuHmzZv2Y79+/Tr69OmDs2fPlonp+PHjJcfbuXNnWCwWnD9/HsDtMS8//fQTTCaTy2UvLTIyEoMGDcLXX39t7061WCxYuXIlBg8ejNDQ0HJfu2HDBhQXF+Oll16CVnu7Gh43bhwiIiLs57mrHN+roqIiXLt2zT4wvjrvV2WGDh2KyMhI++PExEQAwJNPPikZ5J2YmIji4mL7e1bVczs4OBiBgYHYsmWL0y5eV7blyPH8K8+3336LyMhI9OrVS7Lddu3aISwszL5dd59X5BnsxiLZzZs3D3fffTcCAgIQGxuLZs2a2Sv88+fPQ6vVlrnqJC4uDlFRUfYvLsD2ZVaSGGzfvh3t27dH+/btER0dje3btyM2NhaHDx/GiBEj7K85e/YsTp48Kem+cVQyGLREo0aNqnxcWq0WEyZMwIQJE3D9+nXs3LkTCxcuxC+//ILhw4fbu9RK3HXXXZLHYWFhqFOnjn0sQknCVpKIlRYREQEA+O233wBA0pVSHeUd608//YR//vOfOHTokGRckyuJVInz58+jTp06CAsLkyxv1qyZ5HFqaiqEEHjzzTfx5ptvOt3WlStXULduXfvj0olejRo1AMD+Jdm1a1cMGTIESUlJmDNnDrp164bBgwdjxIgRMBgMLh8LADz11FNYuXIltm/fji5dumDDhg3IysrCqFGjKnxdyXlc+rgDAwPRuHFjyXnuiuzsbCQlJWHFihVlzuXc3NxqbbMipWNekvgkJCQ4XV7yXlT13DYYDHjvvffwt7/9DbGxsXjggQfw5z//GU899RTi4uJc2laJgIAA1KtXr9JjO3v2LHJzcyXj7RyVxNcT5xW5H5Mdkl2HDh3sV2OVpypfpJ06dcKiRYvw22+/Yfv27ejcuTM0Gg06deqE7du3Iz4+Hlar1T7+ALCN2WnZsiU++OADp9ssXUk7/lJ2Rc2aNTFw4EAMHDgQ3bp1w9atW3H+/Hn72J6qKLkM9ssvv7RX7I4cfzm7g7Nj3b59OwYOHIguXbpg/vz5qFOnDvR6PZYsWYLly5e7df+OSo795ZdfLnegb+mEWKfTOV2vpNVFo9Hgu+++w+7du/Hjjz9i3bp1GDt2LN5//33s3r27TAJWFX369EFsbCy++uordOnSBV999RXi4uLQs2dPl7flDsOGDcOuXbvwyiuvoHXr1ggLC4PVakXfvn09cll1eTGv7L1w5dx+6aWXMGDAAKxevRrr1q3Dm2++iZkzZ2LTpk1o06aNy58Tg8EgaU0rj9VqRUxMDJYtW+b0+ZIfTJ44r8j9mOyQojRo0ABWqxVnz55FixYt7MuzsrKQk5MjSRZKkpjk5GTs3bsXr7/+OgDbYOQFCxYgPj4eoaGhaNeunf01TZo0weHDh9GjR49qtUxUR/v27bF161ZcvnxZUv6zZ8+ie/fu9sf5+fm4fPky+vfvby8rAMTExFT45dm4cWMAwLFjxyosR3WO9z//+Q+CgoKwbt06ya/UJUuWuLwtwPb+bty4Efn5+ZIvgdOnT0vWKzkmvV7v9sThgQcewAMPPIC3334by5cvx8iRI7FixQr89a9/dbp+RXHT6XQYMWIEli5divfeew+rV6/GuHHjyv2yL1FyHpw+fdp+rABQXFyMc+fOVeuYb9y4gY0bNyIpKQlvvfWWfXlJy0dl5Po8AFU/tx3X/9vf/oa//e1vOHv2LFq3bo33338fX331lcvbcqWMGzZswEMPPVSlHz2unlckL47ZIUUp+aKfO3euZHlJS4zj1T+NGjVC3bp1MWfOHJhMJvtl3507d0ZaWhq+++47PPDAA5JfdsOGDcPvv/+ORYsWldn3rVu3UFBQUK1yZ2Zm4sSJE2WWFxcXY+PGjU675j799FNJH/+CBQtgNpvRr18/ALZWg4iICLzzzjtOxwKUXFZdu3ZtdOnSBYsXL8aFCxck6wiHS/NLxpC4MhOwTqeDRqOBxWKxL0tPTy9z1VpV9e/fH2azWTIVgMViwccffyxZLyYmBt26dcO///1vXL58ucx2qjNv0Y0bNyTxAGCfPK/0tAOOKovbqFGjcOPGDTz77LPIz8/Hk08+WWlZevbsicDAQHz00UeSMn322WfIzc11+So34HZrSuljLP1ZKk91zo/qquq5XVhYiKKiIslzTZo0QXh4uP09q+q2XDVs2DBYLBbMmDGjzHNms9kep+qeVyQvtuyQorRq1QqjR4/Gp59+ipycHHTt2hW//vorPv/8cwwePFjSEgLYEpsVK1agZcuW9jEabdu2RWhoKM6cOSMZrwPYvpi++eYbPPfcc9i8eTMeeughWCwWnDp1Ct988w3WrVtXaRebMxkZGejQoQMefvhh9OjRA3Fxcbhy5Qq+/vprHD58GC+99JL9suUSxcXF6NGjB4YNG4bTp09j/vz56NSpEwYOHAjANtZgwYIFGDVqFNq2bYvhw4ejdu3auHDhAn7++Wc89NBD+OSTTwAAH330ETp16oS2bdti/PjxaNSoEdLT0/Hzzz/bp/8vaeH6xz/+geHDh0Ov12PAgAEVDqR95JFH8MEHH6Bv374YMWIErly5gnnz5qFp06Y4cuSIy3EaMGAAHnroIbz++utIT0/HPffcg++//97peJJ58+ahU6dOaNmyJcaNG4fGjRsjKysLKSkpyMjIwOHDh13a9+eff4758+fj0UcfRZMmTXDz5k0sWrQIERER9iTbmcri1qZNG9x33332we9t27attCy1a9fGlClTkJSUhL59+2LgwIH2c+BPf/pTlRKm0iIiItClSxfMmjULJpMJdevWxfr16+3zV1WmSZMmiIqKwsKFCxEeHo7Q0FAkJia6NG7NlbJW5dw+c+aM/TNyzz33ICAgAKtWrUJWVhaGDx/u0rZc1bVrVzz77LOYOXMmDh06hN69e0Ov1+Ps2bP49ttv8eGHH+Lxxx+v9nlFMvPadWDkd0ouEa7sEk2TySSSkpJEo0aNhF6vFwkJCWLKlCmSS5BLzJs3TwAQzz//vGR5z549BQCxcePGMq8pLi4W7733nrj33nuFwWAQNWrUEO3atRNJSUkiNzfXvh4AMWHChCodW15envjwww9Fnz59RL169YRerxfh4eGiY8eOYtGiRZJLokvisHXrVjF+/HhRo0YNERYWJkaOHCmuX79eZtubN28Wffr0EZGRkSIoKEg0adJEjBkzRuzbt0+y3rFjx8Sjjz4qoqKiRFBQkGjWrJl48803JevMmDFD1K1bV2i1Wsnl1BUd62effSbuuusuYTAYRPPmzcWSJUvsl0M7qsql50IIcf36dTFq1CgREREhIiMjxahRo8TBgwedXvaclpYmnnrqKREXFyf0er2oW7eu+POf/yy+++67MvEsfV5t3rxZABCbN28WQghx4MAB8cQTT4j69esLg8EgYmJixJ///OcycUSpS88riluJkmkD3nnnnUqP39Enn3wimjdvLvR6vYiNjRXPP/+8uHHjhmQdVy49z8jIsJ8DkZGRYujQoeLSpUtOj8mZH374Qdxzzz0iICBA8n6Ud+l56UvCS2L+7bffSpZX9B5VdG5fu3ZNTJgwQTRv3lyEhoaKyMhIkZiYKL755psyZa/K52T06NEiNDTU6bGXPsYSn376qWjXrp0IDg4W4eHhomXLluLVV18Vly5dEkJU/bwi79IIUar9jYg8qmSSsr1791arFYmU58MPP8SkSZOQnp5e5golIvI+jtkhIroDQgh89tln6Nq1KxMdIoXimB0iomooKCjAf//7X2zevBlHjx7FDz/84O0iEVE5mOwQEVXD1atXMWLECERFReHvf/+7fWA5ESkPx+wQERGRqnHMDhEREakakx0iIiJSNY7Zge0eKJcuXUJ4eLisU6YTERFR9QkhcPPmTcTHx1d4zzMmOwAuXbpU5gaQRERE5BsuXrxY4d3smewACA8PB2ALVkREhNu2azKZsH79evs04+QZjLN8GGt5MM7yYJzl4ck45+XlISEhwf49Xh4mO7h9t9+IiAi3JzshISGIiIjgB8mDGGf5MNbyYJzlwTjLQ444VzYEhQOUiYiISNWY7BAREZGqMdkhIiIiVWOyQ0RERKrGZIeIiIhUzavJzrZt2zBgwADEx8dDo9Fg9erVkuc1Go3Tf//617/s6zRs2LDM8++++67MR0JERERK5dVkp6CgAK1atcK8efOcPn/58mXJv8WLF0Oj0WDIkCGS9aZPny5Z74UXXpCj+EREROQDvDrPTr9+/dCvX79yn4+Li5M8/uGHH9C9e3c0btxYsjw8PLzMukRERESAD00qmJWVhZ9//hmff/55mefeffddzJgxA/Xr18eIESMwadIkBASUf2hGoxFGo9H+OC8vD4Bt4iOTyeS2Mpdsy53bpLIYZ/kw1vJgnOXBOMvDk3Gu6jZ9Jtn5/PPPER4ejscee0yy/MUXX0Tbtm0RHR2NXbt2YcqUKbh8+TI++OCDcrc1c+ZMJCUllVm+fv16hISEuL3sycnJbt8mlcU4y4exlgfjLA/GWR6eiHNhYWGV1tMIIYTb914NGo0Gq1atwuDBg50+37x5c/Tq1Qsff/xxhdtZvHgxnn32WeTn58NgMDhdx1nLTkJCAq5du+b220UkJyejV69enIrcgxhn+TDW8mCc5cE4y8OTcc7Ly0OtWrWQm5tb4fe3T7TsbN++HadPn8bKlSsrXTcxMRFmsxnp6elo1qyZ03UMBoPTREiv13vkhPfUdkmKcZYPYy0PxlkejLM8PBHnqm7PJ+bZ+eyzz9CuXTu0atWq0nUPHToErVaLmJgYGUpGREREzhSZLLBYFdF55N2Wnfz8fKSmptofnzt3DocOHUJ0dDTq168PwNZE9e233+L9998v8/qUlBTs2bMH3bt3R3h4OFJSUjBp0iQ8+eSTqFGjhmzHQURERLcVGM1oMz0ZTWLC8N//ecDbxfFusrNv3z50797d/njy5MkAgNGjR2Pp0qUAgBUrVkAIgSeeeKLM6w0GA1asWIFp06bBaDSiUaNGmDRpkn07REREJL9952+g2GLFyct53i4KAC8nO926dUNl46PHjx+P8ePHO32ubdu22L17tyeKRkRERCrhE2N2iIiIiKqLyQ4RERGpGpMdIiIiUjUmO0RERKRqTHaIiIhI1ZjsEBERkaox2SEiIiJVY7JDREREqsZkh4iIiNyqsgmD5cZkh4iIiFSNyQ4RERG5lUaj8XYRJJjsEBERkVuxG4uIiIhIRkx2iIiIyK3YjUVERESqxm4sIiIiIhkx2SEiIiK3YjcWERERqRq7sYiIiIhkxGSHiIiIVI3JDhEREakakx0iIiJSNSY7REREpGpMdoiIiEjVmOwQERGRqjHZISIiIlVjskNERESqxmSHiIiIVI3JDhEREakakx0iIiJSNSY7REREpGpMdoiIiEjVmOwQERGRqjHZISIiIlVjskNERESqxmSHiIiIVI3JDhEREamaV5Odbdu2YcCAAYiPj4dGo8Hq1aslz48ZMwYajUbyr2/fvpJ1srOzMXLkSERERCAqKgrPPPMM8vPzZTwKIiIiUjKvJjsFBQVo1aoV5s2bV+46ffv2xeXLl+3/vv76a8nzI0eOxPHjx5GcnIyffvoJ27Ztw/jx4z1ddCIiIvIRAd7ceb9+/dCvX78K1zEYDIiLi3P63MmTJ7F27Vrs3bsX7du3BwB8/PHH6N+/P2bPno34+Hi3l5mIiIiqTgjh7SIof8zOli1bEBMTg2bNmuH555/H9evX7c+lpKQgKirKnugAQM+ePaHVarFnzx5vFJeIiIgUxqstO5Xp27cvHnvsMTRq1AhpaWn4+9//jn79+iElJQU6nQ6ZmZmIiYmRvCYgIADR0dHIzMwsd7tGoxFGo9H+OC8vDwBgMplgMpncVv6Sbblzm1QW4ywfxloejLM8GGfPsVgs9r89GeeqblPRyc7w4cPtf7ds2RL3338/mjRpgi1btqBHjx7V3u7MmTORlJRUZvn69esREhJS7e2WJzk52e3bpLIYZ/kw1vJgnOXBOLvfyRsaADoAwPrkDdBqPBPnwsLCKq2n6GSntMaNG6NWrVpITU1Fjx49EBcXhytXrkjWMZvNyM7OLnecDwBMmTIFkydPtj/Oy8tDQkICevfujYiICLeV12QyITk5Gb169YJer3fbdkmKcZYPYy0PxlkejLPnhJ29hoWnDgAAevXqiY0bNngkziU9M5XxqWQnIyMD169fR506dQAAHTt2RE5ODvbv34927doBADZt2gSr1YrExMRyt2MwGGAwGMos1+v1HjnhPbVdkmKc5cNYy4Nxlgfj7H46nc7+tz7AFltPxLmq2/NqspOfn4/U1FT743PnzuHQoUOIjo5GdHQ0kpKSMGTIEMTFxSEtLQ2vvvoqmjZtij59+gAAWrRogb59+2LcuHFYuHAhTCYTJk6ciOHDh/NKLCIiIgXw/rVYXr4aa9++fWjTpg3atGkDAJg8eTLatGmDt956CzqdDkeOHMHAgQNx991345lnnkG7du2wfft2SavMsmXL0Lx5c/To0QP9+/dHp06d8Omnn3rrkIiIiEhhvNqy061btwqvv1+3bl2l24iOjsby5cvdWSwiIiJSEcXPs0NERES+i5MKEhERkep4P72RYrJDREREqsZkh4iIiNxK4/C3Elp5mOwQERGRWykhwXHEZIeIiIhUjckOERERuZWkG0sBzTxMdoiIiMitFJDfSDDZISIiIlVjskNERERuxauxiIiISNWUkOA4YrJDREREqsZkh4iIiDxHAZdjMdkhIiIiVWOyQ0RERKrGZIeIiIg8xvudWEx2iIiISOWY7BAREZHHKGB8MpMdIiIiUjcmO0RERKRqTHaIiIjIY4QChigz2SEiIiJVY7JDREREqsZkh4iIiDyGV2MREREReRiTHSIiIlI1JjtERETkMQroxWKyQ0REROrGZIeIiIhUjckOEREReQyvxiIiIiLyMCY7REREpGpMdoiIiMiDvN+PxWSHiIiIVI3JDhEREakakx0iIiLyGF6NRURERORhTHaIiIhI1bya7Gzbtg0DBgxAfHw8NBoNVq9ebX/OZDLhtddeQ8uWLREaGor4+Hg89dRTuHTpkmQbDRs2hEajkfx79913ZT4SIiIickYBvVjeTXYKCgrQqlUrzJs3r8xzhYWFOHDgAN58800cOHAA33//PU6fPo2BAweWWXf69Om4fPmy/d8LL7wgR/GJiIjIBwR4c+f9+vVDv379nD4XGRmJ5ORkybJPPvkEHTp0wIULF1C/fn378vDwcMTFxXm0rEREROSbvJrsuCo3NxcajQZRUVGS5e+++y5mzJiB+vXrY8SIEZg0aRICAso/NKPRCKPRaH+cl5cHwNZ1ZjKZ3Fbekm25c5tUFuMsH8ZaHoyzPBhnz7GYzfa/iz0Y56puUyOEEi4KAzQaDVatWoXBgwc7fb6oqAgPPfQQmjdvjmXLltmXf/DBB2jbti2io6Oxa9cuTJkyBU8//TQ++OCDcvc1bdo0JCUllVm+fPlyhISE3PGxEBER+bPjNzT49JQOAPBOezNC9Z7ZT2FhIUaMGIHc3FxERESUu55PJDsmkwlDhgxBRkYGtmzZUuEBLV68GM8++yzy8/NhMBicruOsZSchIQHXrl2rcNuuMplMSE5ORq9evaDXe+idJsZZRoy1PBhneTDOnrPlzFWM+/IgAGDXK52wd8cWj8Q5Ly8PtWrVqjTZUXw3lslkwrBhw3D+/Hls2rSp0mQkMTERZrMZ6enpaNasmdN1DAaD00RIr9d75IT31HZJinGWuphdiGKLFU1qh7l924y1PBhneTDO7uc4lKTkb0/EuarbU3SyU5LonD17Fps3b0bNmjUrfc2hQ4eg1WoRExMjQwmJlEkIgc6zNgMAjkzrjYggVuREJCNF9Bnd5tVkJz8/H6mpqfbH586dw6FDhxAdHY06derg8ccfx4EDB/DTTz/BYrEgMzMTABAdHY3AwECkpKRgz5496N69O8LDw5GSkoJJkybhySefRI0aNbx1WESKciWviMkOEfk1ryY7+/btQ/fu3e2PJ0+eDAAYPXo0pk2bhv/+978AgNatW0tet3nzZnTr1g0GgwErVqzAtGnTYDQa0ahRI0yaNMm+HSJ/pYyReETktzS3/1RCfeTVZKdbt26oaHx0ZWOn27Zti927d7u7WERERHQnFJDgOOK9sYhUSGH1DBGRVzHZISIiIvdy7MbyXinsmOwQqZBjF7AS+suJyM8orN5hskNERESqxmSHiIiIPEcBzctMdohUyPtVCxGRcjDZISIiIlVjskOkQgpoNSYiPyYc2peVUB0x2SEiIiJVY7JDpEJCEb+liIiUgckOERERuZVjV7oSutWZ7BCpkBIqFyIipWCyQ0RERKrGZIeIiIjcStKN5b1i2DHZISIi8hNbTl/BmqOXvV0M2QV4uwBE5H4cs0NEzoxZshcA8OvfeyAmIshj+1FaFcSWHSIiIj9zvaBYtn0JBfz6YrJDpEKcZ4eISnNMOqwKSEDkxGSHiIjID8g5940SWnMcMdkhUiGF1TNEpDAWq3yVhBKqIyY7REREfsAx6WA3FhH5PP+qxojIVZ5OdpRWBzHZISIi8gPSAcpy7le+fZWHyQ6RCiltcCARKYucY3aUgMkOERGRH5BzzI7Sfm8x2SEiIvIDcl56rjRMdohUyM/qMSJykee7sZRVCzHZISIi8gOOM6vz0nMi8nl+Vo8RkYvkTHaUcMEEkx0iIiI/4JhzWK3y7UsJmOwQqZHCKhoiUhZ2YxEREZGqydqNJdueysdkh0iFhCKqFyJSEkk3lqfveu7ZzbuMyQ4REZGf4QzKROTz/Kw7noiqQM5Lz5U2gSGTHSKVU0A9Q0QKwwHKROTz/KsaI6KqkPPSc6VhskNEROQHZL0RqMPelHDBhFeTnW3btmHAgAGIj4+HRqPB6tWrJc8LIfDWW2+hTp06CA4ORs+ePXH27FnJOtnZ2Rg5ciQiIiIQFRWFZ555Bvn5+TIeBZHyKGHGUiJSLnZjyaigoACtWrXCvHnznD4/a9YsfPTRR1i4cCH27NmD0NBQ9OnTB0VFRfZ1Ro4ciePHjyM5ORk//fQTtm3bhvHjx8t1CERERD7B8UeQnBdjKSGvCvDmzvv164d+/fo5fU4Igblz5+KNN97AoEGDAABffPEFYmNjsXr1agwfPhwnT57E2rVrsXfvXrRv3x4A8PHHH6N///6YPXs24uPjZTsWIiVRQN1CRArm6UvPlZDgOPJqslORc+fOITMzEz179rQvi4yMRGJiIlJSUjB8+HCkpKQgKirKnugAQM+ePaHVarFnzx48+uijTrdtNBphNBrtj/Py8gAAJpMJJpPJbcdQsi13bpPKYpzLMjvEwp3nNWMtD8ZZHv4WZ8fjNJvNHj1ui8Ui2Vfp/btLVbep2GQnMzMTABAbGytZHhsba38uMzMTMTExkucDAgIQHR1tX8eZmTNnIikpqczy9evXIyQk5E6LXkZycrLbt0llMc635RUDJR/vbdu3I9XNpzVjLQ/GWR7+EudbZqCkXjh67DjWXD/msX0dvKYBoAMA7NixA7WCPBPnwsLCKq2n2GTHk6ZMmYLJkyfbH+fl5SEhIQG9e/dGRESE2/ZjMpmQnJyMXr16Qa/Xu227JMU4l3Ut34g3928FAHTu1BnN4sLdsl3GWh6Mszz8Lc55t0x4fe9mAEDze+5B/44NPLYv65HL+PzsUQDAQw91wun9OzwS55KemcooNtmJi4sDAGRlZaFOnTr25VlZWWjdurV9nStXrkheZzabkZ2dbX+9MwaDAQaDocxyvV7vkRPeU9slKcb5Nl3A7SbkAH2A2+PCWMuDcZaHv8Q5wKHHR6PRevSYdQG304sAve1vT8S5qttT7Dw7jRo1QlxcHDZu3GhflpeXhz179qBjx44AgI4dOyInJwf79++3r7Np0yZYrVYkJibKXmYiJVLaQEEi8j7P3y7C8X4RHt1VlXi1ZSc/Px+pqan2x+fOncOhQ4cQHR2N+vXr46WXXsI///lP3HXXXWjUqBHefPNNxMfHY/DgwQCAFi1aoG/fvhg3bhwWLlwIk8mEiRMnYvjw4bwSi/ybAioXIlIW6b2xvFgQL/BqsrNv3z50797d/rhkHM3o0aOxdOlSvPrqqygoKMD48eORk5ODTp06Ye3atQgKCrK/ZtmyZZg4cSJ69OgBrVaLIUOG4KOPPpL9WIiUii07RFSav9313KvJTrdu3Sqc6VWj0WD69OmYPn16uetER0dj+fLlnigeERGRakjvRC5fsuP3t4sgIs8Qkr+9X9EQkfdJ743ltWJ4BZMdIiIiP+Nv3VhMdohUSNpc7b1yEJFyOHZdebobS2l1EJMdIiIiP2NRQgYiIyY7RCrEcTpEVBrH7BAREZHf8HTDjuMPLiU0IjHZIVIhpfWXE5H3SSc19q+KgckOERGRH5AzwVHajywmO0QqxHl2iKhCMlYLSqiBmOwQERH5A2Xdm1NWLic7BQUFeOutt3DfffchLCwM4eHhuP/++zF9+nQUFhZ6ooxE5CLpfBpeLAgRKZKc8+wogUv3xiouLkbXrl1x7Ngx9OvXDwMGDIAQAidPnsTbb7+NX375Bdu2bYNer/dUeYmIiKgavJV/yHkfrvK4lOwsWLAAGRkZOHz4MJo1ayZ57tSpU+jWrRsWLlyIF154wa2FJCLXCD9uriYi5/z5Kk2XurG+//57vPnmm2USHQBo3rw5/vGPf+C7775zW+GIiIjI/Tyd6ygtl3Ip2Tlx4gS6detW7vPdu3fHiRMn7rRMRORGSmhCJiLv89ZEf0qogVxKdnJyclCzZs1yn69ZsyZyc3PvuFBERERE7uJSsmO1WqHT6crfmFYLi8Vyx4UiojvDMTtEVJqcMygrrUXZpQHKQgj06NEDAQHOX2Y2m91SKCIiIvIcWXMRBeQ9LiU7U6dOrXSdIUOGVLswROQeSrsJHxF5n5/lNxJuT3aIiIhIeZTWtSQnl5Kd8mzduhUFBQXo2LEjatSo4Y5NEpHb+G8FR0TOyZn4KOH+fC4lO++99x7y8/MxY8YMALZg9evXD+vXrwcAxMTEYOPGjbj33nvdX1IiqjI//gFHROWQ9cIFhdVBLl2NtXLlStx33332x9999x22bduG7du349q1a2jfvj2SkpLcXkgiqj4mPkTk71xKds6dO4f777/f/njNmjV4/PHH8dBDDyE6OhpvvPEGUlJS3F5IInIN8xsiqoiskwoqoEJyKdkxm80wGAz2xykpKXjwwQftj+Pj43Ht2jX3lY6I7pgC6hkiUgBZ59lRWM3jUrLTpEkTbNu2DQBw4cIFnDlzBl26dLE/n5GRUeEMy0QkD3++6oKIKudvVYRLA5QnTJiAiRMnYvv27UhJScEDDzyAe+65x/78pk2b0KZNG7cXkoiqz98qNSJyzlutLUqoglxq2Rk3bhw+/vhjZGdno1u3bli1apXk+UuXLmHs2LFuLSARuU4JlQsRKZfH73qusErI5XtjXblyBVlZWdi7dy/mzJmDW7du2Z+fP38+Hn30UbcXkoiqj11aRASUGrPjZ9WCS8nO22+/jb///e8IDw9H3bp18eGHH2LChAmeKhsRVZO/VWRE5CpPD1B2+FsB9ZFLyc4XX3yB+fPnY926dVi9ejV+/PFHLFu2DFar1VPlI6I7pIB6hogUwJ/rApeSnQsXLqB///72xz179oRGo8GlS5fcXjAiuhP+XK0RkTOOXdpKaG2Rk8vz7AQFBUmW6fV6mEwmtxaKiNzH3yo1Iqqcp+sFOef0qQqXLj0XQmDMmDGSiQWLiorw3HPPITQ01L7s+++/d18JichlTHCIqDR/rhZcSnZGjx5dZtmTTz7ptsIQkfsp4VcVESmLv9ULLiU7S5Ys8VQ5iMiN/KsaI6KqkPPSc8dkSgktzS6N2SEiH6SAioaIlMXfqgXFJzsNGzaERqMp869kfp9u3bqVee65557zcqmJvEsJv6SISGnkqxiUVge51I3lDXv37oXFYrE/PnbsGHr16oWhQ4fal40bNw7Tp0+3Pw4JCZG1jERKprA6h4i8xJ9nUFZ8slO7dm3J43fffRdNmjRB165d7ctCQkIQFxcnd9GIFMvfBh8SkWv8rY5QfDeWo+LiYnz11VcYO3YsNBqNffmyZctQq1Yt3HfffZgyZQoKCwu9WEoiZfG3X3BE5Jwo94GH96UAim/ZcbR69Wrk5ORgzJgx9mUjRoxAgwYNEB8fjyNHjuC1117D6dOnK5zrx2g0wmg02h/n5eUBAEwmk1snSCzZFidd9CzGuSyTyWz/22wxuy02jLU8GGd5+FuczQ71gtVq9ehxOw4/KamPPLG/qm5TI3zolsh9+vRBYGAgfvzxx3LX2bRpE3r06IHU1FQ0adLE6TrTpk1DUlJSmeXLly/neB9Shd8LgFlHbL9lnm9hQfMon/mYE5GHXCoE3jtsqxfa17Ji1F2eu6/ljkwNvj2nAwC83NKMhDDP7KewsBAjRoxAbm4uIiIiyl3PZ1p2zp8/jw0bNlQ6O3NiYiIAVJjsTJkyBZMnT7Y/zsvLQ0JCAnr37l1hsFxlMpmQnJyMXr16Qa/Xu227JMU4l3Uq8yZmHUkBAPypw5/QuWktt2yXsZYH4ywPf4vz6cybeO+wrV6Ij6+L/v1bemxfN/ZcwLfnTgEAEhMfwKXjuz0S55Kemcr4TLKzZMkSxMTE4JFHHqlwvUOHDgEA6tSpU+46BoNBcsuLEnq93iMnvKe2S1KM82063e2PdoAuwO1xYazlwTjLw1/iHKC/XS9otBqPHrNOp7v9d4Dtb0/Euarb84lkx2q1YsmSJRg9ejQCAm4XOS0tDcuXL0f//v1Rs2ZNHDlyBJMmTUKXLl1w//33e7HERMrBDiwiAvz7YgWfSHY2bNiACxcuYOzYsZLlgYGB2LBhA+bOnYuCggIkJCRgyJAheOONN7xUUiJl8LfLSonINZ6/XYSy+ESy07t3bzgbR52QkICtW7d6oUREvsOHrkEgIg+STCropf16i0/Ns0NEVaOEyoWIlMvTP4KUVgcx2SFSOYXVOUTkJf7cvc1kh4iIyA94rRtLxn2Vh8kOkdopoaYhImXx9ABlhfVjMdkhUiGF1TNERF7FZIdI5fy5n56InJOzXlBCKw+THSIVYoJDRKVJxuz42Tw7THaIVE4BP6qISAEcfwT5W73AZIdIhfytIiMi5VJCdcRkRyVW7r2A+VtSvV0MUiAmPkQElL703L8mFfSJ20VQ5V77z1EAQP/76qBhrVAvl4a8TWH1DBEpjNKSEU9jy47K5BvN3i4CKYyf1WlEVA5Rzt+e3pcSKiEmOypgtd4+kzQaLxaEFEMJl3oSESkFkx0VsDp8sWnAbIekmPgQESCtC+SsFpRQAzHZUQGHhh227BAAZVQuRKQsooJHbt+Xwn5kMdlRAceWHS2zHSpFWVUOESmBwnIRj2OyowKSbizmOgT/q8iIqHLeqheU0MrDZEcFJN1Y3isGKZQC6hkiUhh/qxaY7KiAtGWH6Q4RETnjOEDZv9IdJjsqwEvPqSw5Z9QgIiqfEmogJjsqwG4sIiKqjPR2EfLtSwmY7KgAu7GoNEmlprBKh4i8Q9Le62f1ApMdFXDsxvK3flgiInKd528X4Z0JDMvDZEcFrDI2TZJv4IgdIipNCUmHtzDZUQGrl6YAJyIi3+RvvQBMdlTAKvg73lfcKrZg2Z7zuJx7y6P74ZgdIipNzgRHOhja+5VQgLcLQHfOavV2CaiqZq07hSU701EzNBD73+zl7eIQkR/hAGXyaezG8h1bTl8FAFwvKPbofiR3N1bAryoiIm9isqMCkmTHi+WgyvlbPzkRKYecXUtKa0VisqMCbNnxHXK9PUqraIhIWfytXmCyowJWhQ0EIyIi5VHa3DdyYrKjAmzZ8R1yvT9yTgtPRFSa0r6LmOyogMXKZMdXsOWNiLzGj3sBmOyogNLmM6DyydayI2mu5jlBRPKO5VPadxGTHRVgNxYRESmVEr6XmOyogFUBJxJVjWwfep4TRFSKP4/lY7KjAhyzQ0RELvF0N5bCvouY7KgAZ8ul0jjPDhGVJhnLJ+N3hRK+lxSd7EybNg0ajUbyr3nz5vbni4qKMGHCBNSsWRNhYWEYMmQIsrKyvFhi75DMs+P9c4oqwMHCROQt/lz9KDrZAYB7770Xly9ftv/bsWOH/blJkybhxx9/xLfffoutW7fi0qVLeOyxx7xYWu+QdGN5sRxUOW+8P0r4VUVEyuJviY/i73oeEBCAuLi4Mstzc3Px2WefYfny5Xj44YcBAEuWLEGLFi2we/duPPDAA3IX1Wsk3Vj+dgb7GG9MKkhEBJTq3pZzvwqojxSf7Jw9exbx8fEICgpCx44dMXPmTNSvXx/79++HyWRCz5497es2b94c9evXR0pKSoXJjtFohNFotD/Oy8sDAJhMJphMJreVvWRb7tymM8Vm8+19ms0e35/SyBVnd3BMRj1ZXrPl9jlhNlvcti9firUvY5zl4W9xtjh8V1itVs/WQWaLw9+2/Xpif1XdpqKTncTERCxduhTNmjXD5cuXkZSUhM6dO+PYsWPIzMxEYGAgoqKiJK+JjY1FZmZmhdudOXMmkpKSyixfv349QkJC3HkIAIDk5GS3b9PRyRwNAB0AYNeuXbgc7tHdKZan4+wOt4p0ADQAgDVr1nhsP6cczonDhw8j8NIht27fF2KtBoyzPPwlzidv3K4XbtzI8WgddCbj9r72HziAZpGeiXNhYWGV1lN0stOvXz/73/fffz8SExPRoEEDfPPNNwgODq72dqdMmYLJkyfbH+fl5SEhIQG9e/dGRETEHZXZkclkQnJyMnr16gW9Xu+27ZYWeuYqcPIgAKBjxwfRtn6Ux/alRHLF2R3ePrYVeSZbq2L//v09tp+I1OtYcHI/AOD++1uhf5t4t2zXl2LtyxhnefhbnMPOXsPCUwcAAFE1otC/f6LH9pW+5Tf8fDEVANC2bVsUpO33SJxLemYqo+hkp7SoqCjcfffdSE1NRa9evVBcXIycnBxJ605WVpbTMT6ODAYDDAZDmeV6vd4jJ7yntltCq9PZ/9bpdH7xoXXG03F2N0+WVefhc8LXYu2rGGd5+EucHb8rAI1nv5e0t69/CgiwpRqeiHNVt6f4q7Ec5efnIy0tDXXq1EG7du2g1+uxceNG+/OnT5/GhQsX0LFjRy+WUn5W6+2/FTAOjCrgnauxiIhQ6kag/kXRLTsvv/wyBgwYgAYNGuDSpUuYOnUqdDodnnjiCURGRuKZZ57B5MmTER0djYiICLzwwgvo2LGjX12JBQAW3hvLZ8h3I1AiImVQwveSopOdjIwMPPHEE7h+/Tpq166NTp06Yffu3ahduzYAYM6cOdBqtRgyZAiMRiP69OmD+fPne7nU8uOl575E/veH5wQRAaXm3PJwvaC0WkfRyc6KFSsqfD4oKAjz5s3DvHnzZCqRMln9uGnS18g3zw7PBCIqn7/VED41Zoec441AfQfH7BCRt3jr+0EJs7gz2VEBK28E6jPkanER5T4gIn8l5OvFUtwPbyY7KqC0k4qIiJTN334YM9lRAQsH7VBpklOCJwURlbo3lqdbdhR2nTuTHRWQdmORkvH9ISKSH5MdFZCzH5bujHzz7HDQOhFJCT+ek43JjgpwgLLv4CXhRKQEnq6JhLJ6sZjsqAFnUPYdcr09SqtoiMj7pGN2/KtmYLKjAhyf7EP4BhGRl/hZfiPBZEcFeLsIKo3juIjIm5TWisRkRwWsjjMoe7EcVDm+P0TkPf475IHJjgpYJSm014pBVeCNGZQ5aJ2ISvO3eoHJjgrwaiwiIqqMrK05Cpv/jcmOyvhb06Sv8cqNQHlOEFEp/lYvMNlRASsvPfcZsk0qyBOBiEqRc8SD0mogJjsqwDlVfIc3uhl5ThAR4L0fw0r47cVkRwUk8+wo4awir+NZQEQV8bfvCiY7KsAbgfoOr9QvflapEZFzknvmeXpfCqt2mOyojNJOMJLyxu0iiIjKkLGOUEJ1xGRHBaycaMd3eOHt4RlBRIB//whisqMC0mm5vVYMqgL5BijzRCAiKXmvxlJWHcRkRwU4ZocqwgSYiEqTc4CyEgZDM9lRAenVWN4rh7ulXsnHv7em4VaxxdtFcRv55tmRZz9E5DuEH/8wDvB2AcgNVHq7iJ4fbAUA3Cg04fV+zb1cGvfwzsVY6jkniMg3KK3aYcuOCqi1ZafEgQs3vF0En6PC04CI3Eje+2TJuK9yMNlRATnnTvAGrcbbJXAfb7SyqPGcIKI74+leAKXVO0x2VEDtMyhrNerJdjjPDhF5i1B5L0BFmOyogFXlZ62Kch2vVDAqPz2IqIq8NaZTCVUQkx01UMKZ5EFqatmRi5oGqhOR+3n6R5DSfmQx2VEByTw7CjvB3EHDZOeOqPCUIKJqUOP3Q1Ux2VEBST+sCr/adMx1XObPlRoRKYsS6iMmOyqg9kvP2bJzZ9Q4aJ2IXCdkvJhFaT+8meyogNq7sdR06blcVHgaENEd8udbRjPZURk1nsBs2SEici85fxgroZWHyY4KSFt2vH9SuRtbdohI7YrNVo/vQ9bvB4V9FTHZUQHpAGX14aXnrhMq79okUpNfz2Xj7jd+wfwtqbLtU87WFiXUQUx2VMCq8myHyQ4RqdmU748AAGatPe3R/UjG7Hh6nh3Pbt5lTHZUQHI1luJOsTvHXOfOqPGcIKI742+1gqKTnZkzZ+JPf/oTwsPDERMTg8GDB+P0aWnm261bN2g0Gsm/5557zksl9hZ1d1mwZYeIyA289P2ghK8lRSc7W7duxYQJE7B7924kJyfDZDKhd+/eKCgokKw3btw4XL582f5v1qxZXiqxd1gdxrUp4aRyNw5Qdp0/3/CPiJwTMv4wVtrFMgHeLkBF1q5dK3m8dOlSxMTEYP/+/ejSpYt9eUhICOLi4uQunmKof54dZjtERO6lwi+LCig62SktNzcXABAdHS1ZvmzZMnz11VeIi4vDgAED8OabbyIkJKTc7RiNRhiNRvvjvLw8AIDJZILJZHJbeUu25c5tOmNxaNoxW8we35/cBESFxyRXnN3Nk+U1m823/7ZY3LYvX421r2Gc5aGUODv+SPVsvWCx/20VFderd8rq+L1kNkMLzxxbVbepEUprayqH1WrFwIEDkZOTgx07dtiXf/rpp2jQoAHi4+Nx5MgRvPbaa+jQoQO+//77crc1bdo0JCUllVm+fPnyCpMkpfoqVYu9V209kkMbWdApznNv6YV8YM1FLQbWtyI+1GO7AQD8b4otF0+sbcWIpp6fg0IOJccEAB92NFew5p3Ze1WDr1J1AICB9S3oUdcnPuZEfumdQzpk3bK1YHuyXkjJ0mDFb7Z6ISxA4O0/WSp5RfWtStdiy2Xb99LTd1vQuqZn6qDCwkKMGDECubm5iIiIKHc9n2nZmTBhAo4dOyZJdABg/Pjx9r9btmyJOnXqoEePHkhLS0OTJk2cbmvKlCmYPHmy/XFeXh4SEhLQu3fvCoPlKpPJhOTkZPTq1Qt6vd5t2y1t83dHsffqZQDAvffdh/4dEjy2r2ZvrYdVAFfNQUh5rZvH9iOEwP+mJAMAEhIS0L//veWuK1ec3eF/U9bb/+7fv7/H9lN86BK+Sj0GALi7WXP079LILdv1pVj7MsZZHkqJ84dndwK3bGNRPVkv3NyXAfx2AgCgDwxE//7dq7edIjNCA3XQVjCg8vAvp7Hl8nkAwP333w/8ftgjcS7pmamMTyQ7EydOxE8//YRt27ahXr16Fa6bmJgIAEhNTS032TEYDDAYDGWW6/V6j5zwntquncOYFp1W69F9lVzmfi2/2KP7sThcTx+gq9oxeTzObubJsup0Osnf7t6Xr8XaVzHO8vB2nB2TBrnqheru69y1AnSfvQVd7q6NL8Z2KHc9jfb29U86nQ4CnolzVben6GRHCIEXXngBq1atwpYtW9CoUeW/Tg8dOgQAqFOnjodLpxxWFc4paHbo7+W9sVwnVD73EhG5zh2DVlbsvQAA2Hbm6p1vTEaKTnYmTJiA5cuX44cffkB4eDgyMzMBAJGRkQgODkZaWhqWL1+O/v37o2bNmjhy5AgmTZqELl262JrN/IScs2LKxbFlR6foCRKIiO6MXENnJZeel3ouM7cI564VoGOTmpVtxCcpOtlZsGABANvEgY6WLFmCMWPGIDAwEBs2bMDcuXNRUFCAhIQEDBkyBG+88YYXSus9arwRqGOyowFbdlylxgSYiNyndL3wwMyNAIAV4x/AA43LT3iqWp0oba4vRSc7lX1xJyQkYOvWrTKVRsFU2I3lmOxwUkEiojsnTUCcf1vs/u16hcmOr2IHgQqocVJBs2PLDsfsuEwtLXxEJC+1tqQz2VEBSbLjxXK4k2PLjlq+uNVyHETkm0Q5f3tmX8r6XmKyowJVaZr0NZJkx4vlcCerjAciHbOjlggS0R0Rdz7mwVfrEyY7KiDnl6hcHJMdq49+uEqzqPGNIiKfVN3ayFerYyY7quCjZ18FzJJkx4sFcSNZkzaFXQlBROWTa1xiVaoCdxVFaT0OTHZUwKrCLzY1jtkpneyo5biIyPf4W/3DZEcFpAOU1XECS7qx1HEP0DLdWJ6sa5Q2OJCIvM8NQ3Z8tj5hsqMCSpu8yR0cbxehlgSudNIm11Gp5ZwgIvepbr1QndcpoQpisqMCar/0XK1jdvytGZmInJPtdhFV2I86Z9lhsqM6avn+VOXVWKWOw5NJHG8ESkSlSefZ8Wy9oLQfc0x2VEDtY3YU9pmpNmvpMTsqea+IyPdUuxurGvWWEupwJjsq4DgWRAknlTuYVdiyU7olx7MDlOXZDxH5jvLqAqW1wngCkx0VUGMLgRpbdkp3YxERAd65/1954yIrK0pV589RWm3HZEcFrFU8+XyJGsfslOnGkm3MDhERMP2nE/a/zVZh/76o7uzuVX2ZEuogJjtqoMJLz9XYslPmaixFVAFE5K+ST2RBCCGpm1xpZSqdJN0qtmDIgl34cMNZt5XRXZjsqIAaLz1X45gdb00qqJpskYjcavyX+7H+RNYdtOzcft3E5QfQ4q212H/+BuZsOCOpw5WAyY4KSJIdZZ1f1abKbqzSA5S9UwwiUqmcwmIcycgp9/nAgLJf+TtTr1U7MXGsm386clnynMmsrCtnmOyogJxzJ8jFHyYV9GQSp9YxO0II5N4yebsYRIr08PtbMfCTndiVdq3Mc1arQLHZ+b13So8nrIjjuNCSetpZy1CxRVn3+WGyowJqvBGo5HYRKjkmObux1Oq1/xxBq6T1+PVctreLQqQ42QXFAGxjcUozlpPoAK5dKeq4Zkk1bXKS2JSXWHkLkx01UOGYHX+467kn3yy1zrPzzb4MAMDHm5Q3AJJIKZx95m+ZLPa/g/TSr37Hlh1Xxu+UJEnOkh3HZUqogpjsqIBVhd9sqhyzU+ZGoOo4LiJSFmd1ZtEfyY5ep4EhQCd5zuxCsuO46ZL9mC1lX1NRS5I3MNlRATVejaXGMTulm4o9msOp8BYiRFQ1FSU7QQE66HW3v/qv5Bld+nHpWI+VtAiZSv+Sg/K6sQK8XQC6c9IZLb1XDndS46XnZefZISJyP2c/EItMtuTDoNchUHd7Lp21xzMlsyZX1rIj6fKqoGXHcYCyEqpwtuyogNpvBKoWZWdQ9uDVWJL9eGw3RF5ntljx6bY0HM3I9XZRZLf7t+u4mF1YZrmzuqXIbGvZCQ7UQl/qEvRfjmXa/65ssLKzLi+n3VgmZbXsMNlRGbV8salyzI6X5tlR2uRe7qCSU4LcYOW+i3hnzSkM+GSHt4siq6MZuRj+6W50nrW5zHNOepVQVOy8G6vsa8t+uAqMZsz46QT2n8+WPF/yOXR2mXmRw4DoNccykVtc7i5lwWRHBVQ/ZkdZPxCqrXRrlVzz7Cit75zInY5fyvN2Ebzi4MUb5T7ndMzOHy07QXodArTl3xKidN4ihMC/t6bhsx3nMGRBiqTlx2IVMJotmLnmZJntFBSb7X9vOXMNs4/oyqwjJ47ZUQF1zrOjxpYd7zTtKG1yLyKqXPq1ApitVjSNCXf6fEX3sHLWmFv4R8tOsF5X4Z3NHesps8WKR+fvwtHfb3cRWkqN2Vm2+wI2nrpSZju3ii2Sx3km+e/s7ogtOyogVDlmR1mD29yh7AzKntuX4znBlh0i3yGEgMUq0G32FvT8YBvyjWan65VOHSTfA04qzbxbtu1EBOsrbNlxrKdOXr4pSXSA0q3uAuevFzjdjuO8PkrAZEcF1JIMOHJsjFBPAic9jiKZKgNnE375OrWcE3Tn1Fb/CQEYzbfrhqs3jZW+5laxBXlFt5MiZ63hJbdZiQgOgLaCph3HesrZao71lkUIBOmdd0+ZnAxa9iZ2Y6mA44mttBHw1eXYsqOW8bWlK6DCYnmSHbbsELnOahVY/usFtK1fA/fER3h0X445hVVI72FVXje+Y8LS5V+bJUmRszqzJNmJDNZXmOw47s/Zapdzi26va7Vdyu4LmOyogON5fbPIeZOnr1HjmJ3Sl2d6spnXcU+qbNlRxylBCvbjkUt4Y/UxAED6u4/Itl+rkM4+XN7n1zERKd3646zOXHfcdnl5ZLC+bB+YA8eWHWc/lE5l3pTsx+DkTupK5BulpAo5foneLFLHHaEls3Sq5IvtSqkKqfQAvju1K/Uafjx8qcxypU3bTnQnjv2ei5m/nHSo6zxTQZy8fLPyldzE8QisQkha6Iuq0VpfOtlJvXIT567ZxtaEB+lRwZAdyRCCylqF//zxDqdz7CgRW3ZUIKfw9gQGamnZsTh8gNRyI9CMG7ckjwuL3ftejfi/PQCA++tFSlo+2LJDavLnj23z6RhNVkwbeK/kOSFEhVcpucLxZpnu3G5lSo/ZKf2jyGoV0Go1FX6uS4+XOZOVb/87t7C4wnl29qZnw2i2wBCgQ1EVfihtP3u10nWUgC07Pq7IZEGBw4fhplEdLTtq7MbKuCGd6dSd3ViOgwav5UtbkNQyZsfZZGee3P6NgmKczpTv1z25puQqIWli775zxPFmmXJeWWQVQtIa6/jZnrc5FW1mJCP1Sn6Fn+vSFz9cL7j9g3hQm7oVdj1dyC5E0o8ncDQjF6MX/1ppefedL3++HyVhsuPjsguk01KqpmXHySydvi4rr0jy2JVurGKztcIWrjxJ96VG0iyulnl2nN1s0F2u3CxC+7c3YOoPx+zLus3egj5ztzHh8SGeasUsXc96ki3ZuV03OCYu/1p3Grm3TJi55mSFn+uSwcglsv4YVDysfT00qR1W5q7npS3fc0F1M1Iz2fFxn+04J3mcd0sdLTuOY4/UMmanJBGNCtEDqPrVWDcKitHhnQ2Y+PXBctcpmUMDKPurzmRWRwA9OTbgq5TzyC4oxucp5+3LSr4w3NlMvzc9G+nXnM9LQtXjqdnCbzl0M98okK9etQrpVbXOWpWKzJYKj/V6/u3k7NDFHJy8bJtlukHNUABAoJcGFZu9+MOLyY4PO/Z7rj3ZqfHHF2hekVkVN9G8UXi7cnH32BZvKUl2YsODAJQdsFye/x6+hJxCE34+crncdRxbdgqMZumkgipp2ZFzIKRjpVzRBGyuSL9WgKELU9Bt9hZsdjLjLLmm5F1xvJjBWEkLqCschweUbpV1lStl+mzHOWw7e83+2NkAZbNFVNiKdb3ACCEE/nv4EgbP22mf4fiBxtEAgMAKxux4kjfrIiY7PqbkQ1NstuKLlHT78no1QqDXaWCxCmTe4QdTCRwHXV/MLpRtAj5XOF7hUJnCYjN+z7ENUI6JMAAAFm5NgxACl3Nv2Z+rTOlfRqcy8/Dad0dwxqGrpXSLkVrG7Dh2Y7l7UkGtQ0JTbLZKugG0bkp2Sn5dA8DTS/di06kst2xXCW4VW7D51BVJ94tcHM/vHu9vwcg/BupXZv6WVCzbc6Hc5x1/ZFX18+nMjYJidJy5CVO+P1rmuZ+OXEKn9zYh9crtAcQfbTyLhVvT7I+dteykXS3AvM1pZZaXKDJZ0XZGMl50aA0O1uvQtn4NAN5r2fHmlaGqSXbmzZuHhg0bIigoCImJifj118oHVvma/9v+G+6dug5fpqTj7jd+wTf7MuzPdb6rFurVCAGAcqfvLm3l3gt44tPdksRCKbIdymQVwG9XPdv0v+54Ji5m3x5AfPxSLtYeK78l5WaRCT0/2Ibus7eU+wvLsSvu9f/cruj+1DDa/vf564Xo/N5m9Ju7rUotWI4tXgAwctEerNx3Ea87VKQFxWbJFPOu/JralXoNM3856XJyKYTw+FVzji07nmzluVFYjBsO55+7Jn8s3eC6+mDZaQLcbdp/j+N/lu33eGvvwq1peHrpXoz7Yr9H9yMZy/fH/x2TnYJiC3alXS81hq2s1Cs3MWvtaUz76RTKu7Lb8X0/nVW1cVsfbjiLYQtT8O4vp5BXZEKRyYKV+y4iM68IX/9aNrGauPxgmas0SzuakYPvD2RIPl+lL0JwpnRdcctksV9RJleyM6h1vOQxk507tHLlSkyePBlTp07FgQMH0KpVK/Tp0wdXrqinqTjfaMY/fz6JwmIL3vzheJnnX+xxFxKibcnOG6uPVenX/Gv/OYqU365j/pbyfyFYrBU3l1ZF6pV8tJm+Hh9vPFvl1+T80Uce+0cryK60axWtfkd2nL2GZ7/cjx4fbLVXKI98tAPPfXUAe9Ozy6x/JCMH76w5ZX/smCSVmLnmJFpOW4+hC3fhSEYO/usw/834Lo3t3Y4vrTwEs1Ugr8iMwxdznSaqjq0MJQMlv9x9HmOX7pVcZVGi0GjBztTb8ars/buSV2Q/X0b83x78e+tv+NCF9+pmkQmd3tuMJz+r2i9qx3KVl1QZzRYIIZBxo9BesTseh8nNX975DgP7swuKke0wRsNd4+BulPpRkV7FHyXlMVmsyC0sv2w3i0xYuisda45m4sQfdwY3mi3Yduaq25OfDSdtrVTbzly1J/lWq+sJ8MELN/Dj4Uv28lmtAu+sOYkFf9RRBaV+EBzNyMXaPybLc3Qp5xZSr+Tjte+OSK6CvJ5vxKjP9uCV747Yl13IB6b+eALf7c+wt9TuS8/GL8dub3f5HtsPw2O/5+LbfRexfM8FyQ8Kq1Ugr8iEORvO4Nf0bCzcmoaHZm7CPW+txbu/3K4rruQVYcfZa7BYRZW751cfuoTJ3xzGir0Xq7R+VVRnIsAGNUMqfH5Y+3rocndtybK7YsIkj72Z7Khinp0PPvgA48aNw9NPPw0AWLhwIX7++WcsXrwYr7/+utfKlZVXhOtFwMUbhdDpAmAVtl/Ats+x7f9WIWC12v5vtgqsPZaJ1Cs3MbZTI8SEB8H2+0WDlN+ul7ufx9rWRZBeh8fb1UNK2jX8drUAd7/xC/7RvwUa1QpFSKAOMREGWIXtF3GATiOZcXPPb9eReiUfFqvA7zmFiA414EzmTYQHBeCfP59EZl4Rpg+6VzLoDbANfDNZrPhww1kkRAfjua5NANhO6BsFxagVbkCgToukH4/jRqEJ7yefwaDWdaHR2AYVCgjk3jLh5OU8xEcFIyvPiP3ns3HiUh5u/lGRDGwVj0Xbz+GfP5/E1ZtGXLxRiHYNonHoYg4CdVrsTc+GVQg8EqvBXVfy8dv1W1h98HcMaBWPejVCUCNEj+QTWYgK0aN+dCgu3ihEQo0QBOm1uJBdiK2nr9ortWKzFR9vSkXdqGD7MQ5dmIK7YsIwudfduCs2HPlGMwbP2ymJw8Pvb8WDTWqi731xiAkPQpHJgn9v+w0AsDf9BgZ+Il3fEKBFh0bRWHc8C4cu5tiXP7FoN7Qa4OHmsbiQXYBGtUIxpG09HMm4vc73BzPw67lsHLyQg/K8veak5HFOoQkNX/8ZUSF6DGoVj/YNo2E0WxETbsDGk1n4POU86kYFY0RifftrFmxJQ8fGNWG2WiGE7Zfg7zdu4eTlXBRkanBpRzraNojGnA1nsPs3W0L4e84tzNucir3p2fjz/fFonRAFs9WKOclnEGoIwITuTaHTaHCjsBjHL+Xh401nkZVnxOdjOyA6JBBHf8/F57vS0TQmDBtPZUEDDW6ZLKgVZsDbj94nSSoPX7T92r0nPgKhgQGw/nHzxJLPWOqVfBRbrGhZNxKHLubg022/4cEmtVBktuBidiHGPNgQkcF6BAZoYQjQYdPp2z+Mvt2XgWMONz88dDEHpzNvwioEbpksiAgKgE6rhQa2hCPfaEZwoA5BATr8mp6N9cezMPrBBmgQbRsMmldkQlZeEVaW+rI6kpGLPnO2ITYyCPGRQbiWX4zmceG4UViM6JAAhBQCJy7nQaO1XTlzJc+I1Kv5uHrTCKPZgqMZuTickYuJ3ZsiNsKA4MAANKoVAqPJilnrTkuuIFpz7DJOZuZh6c50nLich5hwA0INARjavh4eaFzT9hmMDEbNsEBcvWnEFynn0aR2GAa1jseF7EJ8kZKOxEa2c9xitdVVAVoNQg0BuHijEMcv3e6i+98VhzCgVR2898tp1K0RjHcebYlbJgt0Gg2uFxhxJc+IuMgghATqcMtkwenMm7iab8TyPRfs49rmbU5FdoFtPpiSLqTr+UY0cfjy3H/+RrlXDPWdu93+98p9F1EzNNDpDwMA+Oh4AIAMLP81w+nzJVJ+u26f4wcA/r7K1poaFaJHjpOk86aTG3h2eGdjhfuoiLNuMGfq1QiutLXIWctOoE6LV/s2w8ebUstczfVEh/qY+VhLNHz953K32b5BNIb9KQGNpvxsHzQe4DA2qHaQqGjiZo/TCB+fsa24uBghISH47rvvMHjwYPvy0aNHIycnBz/88EOZ1xiNRhiNt7/s8/LykJCQgGvXriEiwn33QOk1dzvSr1e/r7cy/e6NhUYD/KN/c8SE21pAlqacx9trTntsn3JqHheOr8a2x2MLd+NCtufiKLezM3pjz7lsPLl4n0f30+/eWKw9kaWaS/eJ5PbXTg2x+tAlXMtXXld/abERBozskIB+98Wi11zbD6z/6doYYUE6zFp3FpN7NsXzXRsDAM5m5aP/J7tQMzQQK8d3AADUrxEMjUYDq1XgPwd/x99Xn7Bve1RiAt76cwu88cNxrNz3OxrXCsFv127/+Ega0AKPt62LwAAt3l17Gp/tPI8J3RqjWWwYXlxpa0X7sKMZvXr1gl6vd+tx5+XloVatWsjNza3w+9vnk51Lly6hbt262LVrFzp27Ghf/uqrr2Lr1q3Ys6ds0/q0adOQlJRUZvny5csRElJxU50rZh3W4UqR7aoBDWz/KclzNZo/lv+xrOT53OI/+lS1AvbkWwBaje3f4IZW/F6ggUUAAxtYoSuVKlsFkPy7Br/l2dYxWjQosgA3Tbe3YRFAkA4otgL5Jg2CdcK+f8DW6nLLIt1wSIBAoVmDsACBfLMGkYECZiug0wB5Jtu6Bp0AxO2+dN0f+9JpbNvTa4StVQfSmBRbgFA9EBUImKxAsA4I1An0rWdFkwgg6xbw7W9aZBs1uG607SvaIGCxArl/7DtMX9JSBhT9UfZwvUCx1RaDksdhAUCuCdBrAKPVtm6gViAkwPZanQYw6IDMW7bX1A0RCNQB14sAs9X2fkUF2uL4eyEQFgDUDAIyCoAagbbX5hQDscHA0EYW5JuBC/kaRAUCWy9r0aqmFQ/H2yL0ewHw0wUtjBYNQgIErhk1sAogQAP8XqhBaIBADQNg0AIFZiDbCITpgew/YhATJGAueS8twDWjBlGBApGBQK0ggccbWXHshgY/nNci33T7/SzZ7i2z7T24ZQECtLbtmawa3LLYzoGiP5bfMgOBWtuxZRuByEDgutF23pScJ2EBtlgLACarBloIBOls72+h2bZOsE7A+sf5VWy9XZ5gnYBWYzsfTVYNdBoBi9CgXqjAlVtAuB7Qa4FCMxATLBChB9LzNSgw2cpnspb6LGmAArMGmj/KYPljf3qt7fgAICrQ9vkyW23/blkAi7h9joYEADUMQF6xLfYBDuetRQAljfEa2NYttgImC2D8Y/s6jbB/NjUA7PdKFEDPula0iBL44qwOvxfaVtL+sX643hbvIJ0t/not7Nu5Zbl9bpfE7ZbFFqeMAlvcIgNh/1zotbbPqIAG4XoBowVICAXS80v+b9tWjUABk9W2LwHb/kuOI1xvqxvyTLbthwVI6xGjxfZ3kA5oX0vgXD5wJleLQK1AsVUDrUYgSGtbx4rb50KEXkD/x/KrRRoYdML+OW1Zw4qbJg3S8zUwaIW9LCX1UP1QgSYRAvuvaaHX2uLbuY4Vh65rcCFfg8hA4MotoEGYgFlocL3IVv72tazILbbFsEmEwO8FttjotQKZhbbPXnig7ZjviRLoFGc7L9NvAnOO2TpB6ocKdKljRbgeSMvT4MA1DQK0tuNPz9egex0r7qthxf5rWqTna9A5zorG4QJbL2tx9IYGoQG2z9H5fA2iDQK961pxJleDFlECbWsJXCoEdmVpcV8NgeTftUgIE8gxAvVCBW6aNDiZo0Gh+XZ92jFWoHG4wI1ioHsd27lrFcAXZ7W4XqTBk3dZUDsIuFoE1A6C5DYR2UZb/RXoZModqwB2X9Fgy2Utog0CgxtYERdi+6xcKgTqhdqe33NFi8ENLWgULn290WI7TqsANl7SoHG4QBMP3Uu1sLAQI0aMYLLjLNmRq2XHZDIhOTnZI9ks3cY4y4exlgfjLA/GWR6ejHNVW3Z8fsxOrVq1oNPpkJUlvYwzKysLcXFxTl9jMBhgMBjKLNfr9R454T21XZJinOXDWMuDcZYH4ywPT8S5qtvz+auxAgMD0a5dO2zceHvgl9VqxcaNGyUtPUREROSffL5lBwAmT56M0aNHo3379ujQoQPmzp2LgoIC+9VZRERE5L9Ukez85S9/wdWrV/HWW28hMzMTrVu3xtq1axEbG+vtohEREZGXqSLZAYCJEydi4sSJ3i4GERERKYzPj9khIiIiqgiTHSIiIlI1JjtERESkakx2iIiISNWY7BAREZGqMdkhIiIiVWOyQ0RERKrGZIeIiIhUjckOERERqZpqZlC+E0IIALZbxbuTyWRCYWEh8vLyeEddD2Kc5cNYy4NxlgfjLA9Pxrnke7vke7w8THYA3Lx5EwCQkJDg5ZIQERGRq27evInIyMhyn9eIytIhP2C1WnHp0iWEh4dDo9G4bbt5eXlISEjAxYsXERER4bbtkhTjLB/GWh6MszwYZ3l4Ms5CCNy8eRPx8fHQassfmcOWHQBarRb16tXz2PYjIiL4QZIB4ywfxloejLM8GGd5eCrOFbXolOAAZSIiIlI1JjtERESkakx2PMhgMGDq1KkwGAzeLoqqMc7yYazlwTjLg3GWhxLizAHKREREpGps2SEiIiJVY7JDREREqsZkh4iIiFSNyQ4RERGpGpOdOzRv3jw0bNgQQUFBSExMxK+//lrh+t9++y2aN2+OoKAgtGzZEmvWrJGppL7NlTgfP34cQ4YMQcOGDaHRaDB37lz5CurjXInzokWL0LlzZ9SoUQM1atRAz549Kz3/6TZXYv3999+jffv2iIqKQmhoKFq3bo0vv/xSxtL6Llfr6BIrVqyARqPB4MGDPVtAlXAlzkuXLoVGo5H8CwoK8mwBBVXbihUrRGBgoFi8eLE4fvy4GDdunIiKihJZWVlO19+5c6fQ6XRi1qxZ4sSJE+KNN94Qer1eHD16VOaS+xZX4/zrr7+Kl19+WXz99dciLi5OzJkzR94C+yhX4zxixAgxb948cfDgQXHy5EkxZswYERkZKTIyMmQuue9xNdabN28W33//vThx4oRITU0Vc+fOFTqdTqxdu1bmkvsWV+Nc4ty5c6Ju3bqic+fOYtCgQfIU1oe5GuclS5aIiIgIcfnyZfu/zMxMj5aRyc4d6NChg5gwYYL9scViEfHx8WLmzJlO1x82bJh45JFHJMsSExPFs88+69Fy+jpX4+yoQYMGTHaq6E7iLIQQZrNZhIeHi88//9xTRVSNO421EEK0adNGvPHGG54onmpUJ85ms1k8+OCD4v/+7//E6NGjmexUgatxXrJkiYiMjJSpdDbsxqqm4uJi7N+/Hz179rQv02q16NmzJ1JSUpy+JiUlRbI+APTp06fc9al6cSbXuSPOhYWFMJlMiI6O9lQxVeFOYy2EwMaNG3H69Gl06dLFk0X1adWN8/Tp0xETE4NnnnlGjmL6vOrGOT8/Hw0aNEBCQgIGDRqE48ePe7ScTHaq6dq1a7BYLIiNjZUsj42NRWZmptPXZGZmurQ+VS/O5Dp3xPm1115DfHx8mYSepKob69zcXISFhSEwMBCPPPIIPv74Y/Tq1cvTxfVZ1Ynzjh078Nlnn2HRokVyFFEVqhPnZs2aYfHixfjhhx/w1VdfwWq14sEHH0RGRobHysm7nhPRHXv33XexYsUKbNmyxfMDDf1UeHg4Dh06hPz8fGzcuBGTJ09G48aN0a1bN28XTRVu3ryJUaNGYdGiRahVq5a3i6NqHTt2RMeOHe2PH3zwQbRo0QL//ve/MWPGDI/sk8lONdWqVQs6nQ5ZWVmS5VlZWYiLi3P6mri4OJfWp+rFmVx3J3GePXs23n33XWzYsAH333+/J4upCtWNtVarRdOmTQEArVu3xsmTJzFz5kwmO+VwNc5paWlIT0/HgAED7MusVisAICAgAKdPn0aTJk08W2gf5I46Wq/Xo02bNkhNTfVEEQGwG6vaAgMD0a5dO2zcuNG+zGq1YuPGjZKM1VHHjh0l6wNAcnJyuetT9eJMrqtunGfNmoUZM2Zg7dq1aN++vRxF9XnuOqetViuMRqMniqgKrsa5efPmOHr0KA4dOmT/N3DgQHTv3h2HDh1CQkKCnMX3Ge44ny0WC44ePYo6dep4qpi89PxOrFixQhgMBrF06VJx4sQJMX78eBEVFWW/hG7UqFHi9ddft6+/c+dOERAQIGbPni1Onjwppk6dykvPq8DVOBuNRnHw4EFx8OBBUadOHfHyyy+LgwcPirNnz3rrEHyCq3F+9913RWBgoPjuu+8kl5DevHnTW4fgM1yN9TvvvCPWr18v0tLSxIkTJ8Ts2bNFQECAWLRokbcOwSe4GufSeDVW1bga56SkJLFu3TqRlpYm9u/fL4YPHy6CgoLE8ePHPVZGJjt36OOPPxb169cXgYGBokOHDmL37t3257p27SpGjx4tWf+bb74Rd999twgMDBT33nuv+Pnnn2UusW9yJc7nzp0TAMr869q1q/wF9zGuxLlBgwZO4zx16lT5C+6DXIn1P/7xD9G0aVMRFBQkatSoITp27ChWrFjhhVL7HlfraEdMdqrOlTi/9NJL9nVjY2NF//79xYEDBzxaPo0QQniu3YiIiIjIuzhmh4iIiFSNyQ4RERGpGpMdIiIiUjUmO0RERKRqTHaIiIhI1ZjsEBERkaox2SEiIiJVY7JDREREqsZkh4g8YsyYMdBoNGX+efJmf0REzvCu50TkMX379sWSJUsky2rXri15XFxcjMDAQDmLRUR+hi07ROQxBoMBcXFxkn89evTAxIkT8dJLL6FWrVro06cPAODYsWPo168fwsLCEBsbi1GjRuHatWv2bRUUFOCpp55CWFgY6tSpg/fffx/dunXDSy+9ZF9Ho9Fg9erVkjJERUVh6dKl9scXL17EsGHDEBUVhejoaAwaNAjp6en258eMGYPBgwdj9uzZqFOnDmrWrIkJEybAZDLZ1zEajXjttdeQkJAAg8GApk2b4rPPPoMQAk2bNsXs2bMlZTh06BBbtYi8iMkOEcnu888/R2BgIHbu3ImFCxciJycHDz/8MNq0aYN9+/Zh7dq1yMrKwrBhw+yveeWVV7B161b88MMPWL9+PbZs2YIDBw64tF+TyYQ+ffogPDwc27dvx86dOxEWFoa+ffuiuLjYvt7mzZuRlpaGzZs34/PPP8fSpUslCdNTTz2Fr7/+Gh999BFOnjyJf//73wgLC4NGo8HYsWPLtGYtWbIEXbp0QdOmTasXMCK6Mx69zSgR+a3Ro0cLnU4nQkND7f8ef/xx0bVrV9GmTRvJujNmzBC9e/eWLLt48aIAIE6fPi1u3rwpAgMDxTfffGN//vr16yI4OFj87//+r30ZALFq1SrJdiIjI8WSJUuEEEJ8+eWXolmzZsJqtdqfNxqNIjg4WKxbt85e7gYNGgiz2WxfZ+jQoeIvf/mLEEKI06dPCwAiOTnZ6XH//vvvQqfTiT179gghhCguLha1atUSS5curULUiMgTOGaHiDyme/fuWLBggf1xaGgonnjiCbRr106y3uHDh7F582aEhYWV2UZaWhpu3bqF4uJiJCYm2pdHR0ejWbNmLpXn8OHDSE1NRXh4uGR5UVER0tLS7I/vvfde6HQ6++M6derg6NGjAGxdUjqdDl27dnW6j/j4eDzyyCNYvHgxOnTogB9//BFGoxFDhw51qaxE5D5MdojIY0JDQ5123YSGhkoe5+fnY8CAAXjvvffKrFunTp0qj3XRaDQQQkiWOY61yc/PR7t27bBs2bIyr3UcOK3X68ts12q1AgCCg4MrLcdf//pXjBo1CnPmzMGSJUvwl7/8BSEhIVU6BiJyPyY7ROR1bdu2xX/+8x80bNgQAQFlq6UmTZpAr9djz549qF+/PgDgxo0bOHPmjKSFpXbt2rh8+bL98dmzZ1FYWCjZz8qVKxETE4OIiIhqlbVly5awWq3YunUrevbs6XSd/v37IzQ0FAsWLMDatWuxbdu2au2LiNyDA5SJyOsmTJiA7OxsPPHEE9i7dy/S0tKwbt06PP3007BYLAgLC8MzzzyDV155BZs2bcKxY8cwZswYaLXSKuzhhx/GJ598goMHD2Lfvn147rnnJK00I0eORK1atTBo0CBs374d586dw5YtW/Diiy8iIyOjSmVt2LAhRo8ejbFjx2L16tX2bXzzzTf2dXQ6HcaMGYMpU6bgrrvuQseOHd0TKCKqFiY7ROR18fHx2LlzJywWC3r37o2WLVvipZdeQlRUlD2h+de//oXOnTtjwIAB6NmzJzp16lRm7M/777+PhIQEdO7cGSNGjMDLL78s6T4KCQnBtm3bUL9+fTz22GNo0aIFnnnmGRQVFbnU0rNgwQI8/vjj+J//+R80b94c48aNQ0FBgWSdZ555BsXFxXj66afvIDJE5A4aUbqDm4jIR3Tr1g2tW7fG3LlzvV2UMrZv344ePXrg4sWLiI2N9XZxiPwax+wQEbmR0WjE1atXMW3aNAwdOpSJDpECsBuLiMiNvv76azRo0AA5OTmYNWuWt4tDRGA3FhEREakcW3aIiIhI1ZjsEBERkaox2SEiIiJVY7JDREREqsZkh4iIiFSNyQ4RERGpGpMdIiIiUjUmO0RERKRqTHaIiIhI1f4fmKcUP2A/OvgAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_psd(sample_ts)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Autocorrelation for different lags"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The autocorrelation function (ACF) is a mathematical tool used to measure the correlation between a time series and its lagged versions. In simpler terms, it quantifies the similarity between a signal and a delayed copy of itself at different time lags. It is commonly used in signal processing, time series analysis, and statistics.\n",
    "\n",
    "Mathematically, the autocorrelation function for a time series $\\{X(t)\\}$ at lag k is defined as:\n",
    "\n",
    "$$ACF(k) = Cov(X(t), X(t-k)) / (Var(X(t)) * Var(X(t-k)))$$\n",
    "\n",
    "Where $Cov(X(t), X(t-k))$ is the covariance between $X(t)$ and $X(t-k)$, and $Var(X(t))$ and $Var(X(t-k))$ are the variances of $X(t)$ and $X(t-k)$, respectively.\n",
    "\n",
    "The ACF provides insights into the temporal patterns and dependencies present in a time series. By examining the ACF plot, you can identify whether the time series exhibits any repeating patterns or significant correlations at different lags. For example:\n",
    "\n",
    "1. If the ACF decays quickly, it suggests that the time series is relatively random and uncorrelated over longer time lags.\n",
    "2. If the ACF has a slow decay or periodic spikes, it indicates the presence of seasonality or repeating patterns in the time series.\n",
    "3. If there is a significant spike in the ACF at a particular lag, it may indicate a strong correlation or relationship between the signal and its past values at that specific time lag.\n",
    "\n",
    "\n",
    "Now, let's discuss the usefulness of the autocorrelation function in the context of accelerometer signals for anomaly detection.\n",
    "\n",
    "Accelerometer signals, which measure the acceleration of an object or device, are often used to track and monitor movement or vibrations.  The ACF can be valuable in this context for the following reasons:\n",
    "\n",
    ">- Revealing Periodic Patterns: Accelerometer data from certain activities or healthy machinery may exhibit periodic patterns. By examining the ACF, you can detect the presence of periodicities and use them to understand normal behavior.\n",
    "\n",
    ">- Characterizing Cyclic Anomalies: Anomalies in accelerometer data might manifest as cyclic or repeating patterns. ACF can help identify these cyclic patterns, allowing for the detection of regular anomalies.\n",
    "\n",
    ">- Lag Analysis: The ACF provides information about the correlation between the current and past values of the accelerometer signal. This analysis can be beneficial in understanding the temporal behavior of the system and detecting deviations from the norm.\n",
    "\n",
    ">- Identifying Unusual Behavior: Sudden changes or unexpected behaviors in accelerometer data might cause spikes or distinct patterns in the ACF. Monitoring the ACF can, therefore, assist in identifying such anomalies.\n",
    "\n",
    ">- Windowed ACF for Local Anomaly Detection: By computing the ACF within smaller windows of the time series, you can perform local anomaly detection. Sudden changes in ACF values within specific time segments might indicate localized anomalies."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Other functions for pre-processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We also planned to introduce other features in addition to those described above:\n",
    ">- Season: extracted from the \"when\" column of the original data set, inidicates the season in which the measurement is taken;\n",
    ">- Elapsed time: again extracted from the \"when\" column and indicates the number of months passed since the first month of observation."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data preparation and feature extraction (high dimensionality)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = preproc(df)\n",
    "test = preproc(test_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Season</th>\n",
       "      <th>elapsed_time</th>\n",
       "      <th>E1A_mag_max_1</th>\n",
       "      <th>E1A_mag_max_2</th>\n",
       "      <th>E1A_mag_max_3</th>\n",
       "      <th>E1A_mag_max_4</th>\n",
       "      <th>E1A_mag_max_5</th>\n",
       "      <th>E1A_mag_max_6</th>\n",
       "      <th>E1A_mag_max_7</th>\n",
       "      <th>E1A_mag_max_8</th>\n",
       "      <th>...</th>\n",
       "      <th>M2A_psd3</th>\n",
       "      <th>M2H_psd1</th>\n",
       "      <th>M2H_psd2</th>\n",
       "      <th>M2H_psd3</th>\n",
       "      <th>M2P_psd1</th>\n",
       "      <th>M2P_psd2</th>\n",
       "      <th>M2P_psd3</th>\n",
       "      <th>M2V_psd1</th>\n",
       "      <th>M2V_psd2</th>\n",
       "      <th>M2V_psd3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>6.806409e+04</td>\n",
       "      <td>35431.749770</td>\n",
       "      <td>59566.427102</td>\n",
       "      <td>126189.729664</td>\n",
       "      <td>12104.302024</td>\n",
       "      <td>3776.425253</td>\n",
       "      <td>1448.342098</td>\n",
       "      <td>9170.837140</td>\n",
       "      <td>...</td>\n",
       "      <td>28.452744</td>\n",
       "      <td>0.014132</td>\n",
       "      <td>0.234603</td>\n",
       "      <td>17.168458</td>\n",
       "      <td>0.014573</td>\n",
       "      <td>0.138001</td>\n",
       "      <td>16.675335</td>\n",
       "      <td>0.001447</td>\n",
       "      <td>0.033303</td>\n",
       "      <td>18.282919</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3</td>\n",
       "      <td>20</td>\n",
       "      <td>1.848728e+05</td>\n",
       "      <td>438472.069986</td>\n",
       "      <td>289591.035635</td>\n",
       "      <td>3308.291517</td>\n",
       "      <td>74843.101888</td>\n",
       "      <td>3554.999695</td>\n",
       "      <td>820.542539</td>\n",
       "      <td>753.311788</td>\n",
       "      <td>...</td>\n",
       "      <td>15.837722</td>\n",
       "      <td>0.030442</td>\n",
       "      <td>0.209081</td>\n",
       "      <td>10.461992</td>\n",
       "      <td>0.020580</td>\n",
       "      <td>0.042926</td>\n",
       "      <td>5.746151</td>\n",
       "      <td>0.006826</td>\n",
       "      <td>0.115922</td>\n",
       "      <td>11.918044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>70</td>\n",
       "      <td>2.024669e+06</td>\n",
       "      <td>120144.625912</td>\n",
       "      <td>112000.320858</td>\n",
       "      <td>9512.858234</td>\n",
       "      <td>2416.209026</td>\n",
       "      <td>1622.871783</td>\n",
       "      <td>191.321927</td>\n",
       "      <td>526.334264</td>\n",
       "      <td>...</td>\n",
       "      <td>17.891314</td>\n",
       "      <td>0.018397</td>\n",
       "      <td>0.215008</td>\n",
       "      <td>14.345343</td>\n",
       "      <td>0.009525</td>\n",
       "      <td>0.027912</td>\n",
       "      <td>19.951229</td>\n",
       "      <td>0.006381</td>\n",
       "      <td>0.123472</td>\n",
       "      <td>18.462069</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>75</td>\n",
       "      <td>1.164122e+04</td>\n",
       "      <td>130062.023086</td>\n",
       "      <td>63264.266551</td>\n",
       "      <td>385363.951843</td>\n",
       "      <td>476220.035156</td>\n",
       "      <td>16840.280143</td>\n",
       "      <td>2786.596621</td>\n",
       "      <td>4081.567101</td>\n",
       "      <td>...</td>\n",
       "      <td>14.283665</td>\n",
       "      <td>0.003277</td>\n",
       "      <td>0.276258</td>\n",
       "      <td>21.956153</td>\n",
       "      <td>0.002684</td>\n",
       "      <td>0.072033</td>\n",
       "      <td>21.260119</td>\n",
       "      <td>0.001413</td>\n",
       "      <td>0.051685</td>\n",
       "      <td>12.570201</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1.802098e+04</td>\n",
       "      <td>3651.080879</td>\n",
       "      <td>7153.671289</td>\n",
       "      <td>312.323450</td>\n",
       "      <td>7273.230036</td>\n",
       "      <td>2188.072980</td>\n",
       "      <td>2503.623617</td>\n",
       "      <td>13942.681254</td>\n",
       "      <td>...</td>\n",
       "      <td>20.518775</td>\n",
       "      <td>0.003907</td>\n",
       "      <td>0.061459</td>\n",
       "      <td>10.524084</td>\n",
       "      <td>0.000808</td>\n",
       "      <td>0.011370</td>\n",
       "      <td>13.955317</td>\n",
       "      <td>0.000887</td>\n",
       "      <td>0.022850</td>\n",
       "      <td>15.773869</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>658</th>\n",
       "      <td>1</td>\n",
       "      <td>26</td>\n",
       "      <td>1.284998e+04</td>\n",
       "      <td>122735.369697</td>\n",
       "      <td>24384.588125</td>\n",
       "      <td>192115.450712</td>\n",
       "      <td>67962.611661</td>\n",
       "      <td>3253.077441</td>\n",
       "      <td>3705.208570</td>\n",
       "      <td>7864.876906</td>\n",
       "      <td>...</td>\n",
       "      <td>15.713019</td>\n",
       "      <td>0.004645</td>\n",
       "      <td>0.179328</td>\n",
       "      <td>8.688547</td>\n",
       "      <td>0.004834</td>\n",
       "      <td>0.081339</td>\n",
       "      <td>13.658737</td>\n",
       "      <td>0.002488</td>\n",
       "      <td>0.038790</td>\n",
       "      <td>13.614371</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>659</th>\n",
       "      <td>1</td>\n",
       "      <td>64</td>\n",
       "      <td>4.150732e+03</td>\n",
       "      <td>128007.904321</td>\n",
       "      <td>20370.578567</td>\n",
       "      <td>8427.957587</td>\n",
       "      <td>155044.450928</td>\n",
       "      <td>33908.927987</td>\n",
       "      <td>682.468216</td>\n",
       "      <td>2403.339912</td>\n",
       "      <td>...</td>\n",
       "      <td>29.274844</td>\n",
       "      <td>0.008489</td>\n",
       "      <td>0.064193</td>\n",
       "      <td>21.850531</td>\n",
       "      <td>0.006127</td>\n",
       "      <td>0.026331</td>\n",
       "      <td>19.125511</td>\n",
       "      <td>0.002126</td>\n",
       "      <td>0.072776</td>\n",
       "      <td>30.545321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>660</th>\n",
       "      <td>1</td>\n",
       "      <td>15</td>\n",
       "      <td>1.028444e+04</td>\n",
       "      <td>111601.814605</td>\n",
       "      <td>197658.285183</td>\n",
       "      <td>343008.269175</td>\n",
       "      <td>227935.699029</td>\n",
       "      <td>74277.886723</td>\n",
       "      <td>4596.550570</td>\n",
       "      <td>24059.226406</td>\n",
       "      <td>...</td>\n",
       "      <td>10.944851</td>\n",
       "      <td>0.029701</td>\n",
       "      <td>0.157739</td>\n",
       "      <td>16.822093</td>\n",
       "      <td>0.019558</td>\n",
       "      <td>0.084415</td>\n",
       "      <td>9.870443</td>\n",
       "      <td>0.005622</td>\n",
       "      <td>0.146952</td>\n",
       "      <td>13.225888</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>661</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>4.084544e+05</td>\n",
       "      <td>302934.847480</td>\n",
       "      <td>10649.072514</td>\n",
       "      <td>18817.172722</td>\n",
       "      <td>68414.811460</td>\n",
       "      <td>127692.209605</td>\n",
       "      <td>13454.285278</td>\n",
       "      <td>16568.799289</td>\n",
       "      <td>...</td>\n",
       "      <td>29.468620</td>\n",
       "      <td>0.009112</td>\n",
       "      <td>0.355042</td>\n",
       "      <td>12.764579</td>\n",
       "      <td>0.006090</td>\n",
       "      <td>0.115611</td>\n",
       "      <td>17.350777</td>\n",
       "      <td>0.002193</td>\n",
       "      <td>0.133941</td>\n",
       "      <td>31.157023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>662</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4.895751e+03</td>\n",
       "      <td>152257.272100</td>\n",
       "      <td>119925.141554</td>\n",
       "      <td>28564.405089</td>\n",
       "      <td>10870.523920</td>\n",
       "      <td>7881.093087</td>\n",
       "      <td>11667.694856</td>\n",
       "      <td>59394.006404</td>\n",
       "      <td>...</td>\n",
       "      <td>19.607810</td>\n",
       "      <td>0.006243</td>\n",
       "      <td>0.117144</td>\n",
       "      <td>15.022703</td>\n",
       "      <td>0.003252</td>\n",
       "      <td>0.018423</td>\n",
       "      <td>10.398349</td>\n",
       "      <td>0.001543</td>\n",
       "      <td>0.023879</td>\n",
       "      <td>15.353616</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>663 rows × 554 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Season  elapsed_time  E1A_mag_max_1  E1A_mag_max_2  E1A_mag_max_3  \\\n",
       "0         2             5   6.806409e+04   35431.749770   59566.427102   \n",
       "1         3            20   1.848728e+05  438472.069986  289591.035635   \n",
       "2         3            70   2.024669e+06  120144.625912  112000.320858   \n",
       "3         1            75   1.164122e+04  130062.023086   63264.266551   \n",
       "4         1             3   1.802098e+04    3651.080879    7153.671289   \n",
       "..      ...           ...            ...            ...            ...   \n",
       "658       1            26   1.284998e+04  122735.369697   24384.588125   \n",
       "659       1            64   4.150732e+03  128007.904321   20370.578567   \n",
       "660       1            15   1.028444e+04  111601.814605  197658.285183   \n",
       "661       4             0   4.084544e+05  302934.847480   10649.072514   \n",
       "662       1             3   4.895751e+03  152257.272100  119925.141554   \n",
       "\n",
       "     E1A_mag_max_4  E1A_mag_max_5  E1A_mag_max_6  E1A_mag_max_7  \\\n",
       "0    126189.729664   12104.302024    3776.425253    1448.342098   \n",
       "1      3308.291517   74843.101888    3554.999695     820.542539   \n",
       "2      9512.858234    2416.209026    1622.871783     191.321927   \n",
       "3    385363.951843  476220.035156   16840.280143    2786.596621   \n",
       "4       312.323450    7273.230036    2188.072980    2503.623617   \n",
       "..             ...            ...            ...            ...   \n",
       "658  192115.450712   67962.611661    3253.077441    3705.208570   \n",
       "659    8427.957587  155044.450928   33908.927987     682.468216   \n",
       "660  343008.269175  227935.699029   74277.886723    4596.550570   \n",
       "661   18817.172722   68414.811460  127692.209605   13454.285278   \n",
       "662   28564.405089   10870.523920    7881.093087   11667.694856   \n",
       "\n",
       "     E1A_mag_max_8  ...   M2A_psd3  M2H_psd1  M2H_psd2   M2H_psd3  M2P_psd1  \\\n",
       "0      9170.837140  ...  28.452744  0.014132  0.234603  17.168458  0.014573   \n",
       "1       753.311788  ...  15.837722  0.030442  0.209081  10.461992  0.020580   \n",
       "2       526.334264  ...  17.891314  0.018397  0.215008  14.345343  0.009525   \n",
       "3      4081.567101  ...  14.283665  0.003277  0.276258  21.956153  0.002684   \n",
       "4     13942.681254  ...  20.518775  0.003907  0.061459  10.524084  0.000808   \n",
       "..             ...  ...        ...       ...       ...        ...       ...   \n",
       "658    7864.876906  ...  15.713019  0.004645  0.179328   8.688547  0.004834   \n",
       "659    2403.339912  ...  29.274844  0.008489  0.064193  21.850531  0.006127   \n",
       "660   24059.226406  ...  10.944851  0.029701  0.157739  16.822093  0.019558   \n",
       "661   16568.799289  ...  29.468620  0.009112  0.355042  12.764579  0.006090   \n",
       "662   59394.006404  ...  19.607810  0.006243  0.117144  15.022703  0.003252   \n",
       "\n",
       "     M2P_psd2   M2P_psd3  M2V_psd1  M2V_psd2   M2V_psd3  \n",
       "0    0.138001  16.675335  0.001447  0.033303  18.282919  \n",
       "1    0.042926   5.746151  0.006826  0.115922  11.918044  \n",
       "2    0.027912  19.951229  0.006381  0.123472  18.462069  \n",
       "3    0.072033  21.260119  0.001413  0.051685  12.570201  \n",
       "4    0.011370  13.955317  0.000887  0.022850  15.773869  \n",
       "..        ...        ...       ...       ...        ...  \n",
       "658  0.081339  13.658737  0.002488  0.038790  13.614371  \n",
       "659  0.026331  19.125511  0.002126  0.072776  30.545321  \n",
       "660  0.084415   9.870443  0.005622  0.146952  13.225888  \n",
       "661  0.115611  17.350777  0.002193  0.133941  31.157023  \n",
       "662  0.018423  10.398349  0.001543  0.023879  15.353616  \n",
       "\n",
       "[663 rows x 554 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Import the dataset of correlations"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We thought of evaluating time series correlations between different measurement points on the compressor. To do this we reused the procedure developed in the second homework as well, only this time we have 3 classes and not 2. We need to build a unique dataset for each class by putting the time series as rows. In this way our new datasets will have as number of rows the absolute frequencies of the class for the length of the time series and as columns 24, which is the number of different points of measuraments. In addition, we constructed 3 other datasets in which the 2 different classes than the reference class are present, changing the re-reference class each time. Then in this way we created the graphs of the differences between the correlations of the reference class versus the correlation of the other two classes put together, in a kind of one vs all approach, so as to see if there are relevant differences in the correlations between two point in that class versus the other. This procedure was done for each class, and once a certain threshold was used, we selected the common variables. To take advantage of the code previously we worked on R and exported the 21 variables, these are the correlations that show a greater difference in one class than the others and could be useful for classification."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "corr_train=pd.read_csv(\"train_a.csv\")\n",
    "corr_test=pd.read_csv(\"test_b.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "train=pd.concat([train,corr_train],axis=1)\n",
    "test=pd.concat([test,corr_test], axis=1)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Scaling the data set"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We note that groups of columns have different sizes so we opt for (robust) standardization by columns in our dataframe in order to make the different features comparable and not get strange values for the parameters of our model. It is particularly useful when working with data that contain outliers or scattered values and it's computed as following:\n",
    "\n",
    "$$\n",
    "X_{\\text {scale }}=\\frac{x_i-x_{\\mathrm{med}}}{x_{75}-x_{25}}\n",
    "$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Season</th>\n",
       "      <th>elapsed_time</th>\n",
       "      <th>E1A_mag_max_1</th>\n",
       "      <th>E1A_mag_max_2</th>\n",
       "      <th>E1A_mag_max_3</th>\n",
       "      <th>E1A_mag_max_4</th>\n",
       "      <th>E1A_mag_max_5</th>\n",
       "      <th>E1A_mag_max_6</th>\n",
       "      <th>E1A_mag_max_7</th>\n",
       "      <th>E1A_mag_max_8</th>\n",
       "      <th>...</th>\n",
       "      <th>cor_E2P_M2P</th>\n",
       "      <th>cor_E3P_E4P</th>\n",
       "      <th>cor_E3P_M1P</th>\n",
       "      <th>cor_E3P_M2P</th>\n",
       "      <th>cor_E3V_E4A</th>\n",
       "      <th>cor_E3V_E4V</th>\n",
       "      <th>cor_E4A_E4H</th>\n",
       "      <th>cor_E4P_M2P</th>\n",
       "      <th>cor_M1A_M2A</th>\n",
       "      <th>cor_M1P_M2P</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.638298</td>\n",
       "      <td>-0.053077</td>\n",
       "      <td>-0.262788</td>\n",
       "      <td>0.047737</td>\n",
       "      <td>-0.242141</td>\n",
       "      <td>-0.169405</td>\n",
       "      <td>-0.221405</td>\n",
       "      <td>-0.215468</td>\n",
       "      <td>-0.232382</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.511122</td>\n",
       "      <td>-0.254089</td>\n",
       "      <td>0.096485</td>\n",
       "      <td>-0.548342</td>\n",
       "      <td>1.769077</td>\n",
       "      <td>0.091947</td>\n",
       "      <td>0.726240</td>\n",
       "      <td>-0.476365</td>\n",
       "      <td>0.001820</td>\n",
       "      <td>-1.332654</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.5</td>\n",
       "      <td>0.680851</td>\n",
       "      <td>-0.072831</td>\n",
       "      <td>-0.266944</td>\n",
       "      <td>-0.349170</td>\n",
       "      <td>-0.152923</td>\n",
       "      <td>0.557963</td>\n",
       "      <td>8.095217</td>\n",
       "      <td>-0.112630</td>\n",
       "      <td>0.505699</td>\n",
       "      <td>...</td>\n",
       "      <td>0.612548</td>\n",
       "      <td>-0.220519</td>\n",
       "      <td>-1.113509</td>\n",
       "      <td>0.697665</td>\n",
       "      <td>2.088551</td>\n",
       "      <td>0.484775</td>\n",
       "      <td>-0.262615</td>\n",
       "      <td>0.257704</td>\n",
       "      <td>-0.460445</td>\n",
       "      <td>-0.224991</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.5</td>\n",
       "      <td>-0.446809</td>\n",
       "      <td>-0.075265</td>\n",
       "      <td>-0.297735</td>\n",
       "      <td>-0.293777</td>\n",
       "      <td>-0.248054</td>\n",
       "      <td>-0.185424</td>\n",
       "      <td>-0.242364</td>\n",
       "      <td>-0.142257</td>\n",
       "      <td>-0.231727</td>\n",
       "      <td>...</td>\n",
       "      <td>0.368787</td>\n",
       "      <td>-1.128125</td>\n",
       "      <td>1.083483</td>\n",
       "      <td>0.035716</td>\n",
       "      <td>-1.100787</td>\n",
       "      <td>0.939328</td>\n",
       "      <td>-0.343666</td>\n",
       "      <td>-0.026715</td>\n",
       "      <td>1.450220</td>\n",
       "      <td>1.111149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.510638</td>\n",
       "      <td>1.175205</td>\n",
       "      <td>5.244658</td>\n",
       "      <td>1.275429</td>\n",
       "      <td>-0.107898</td>\n",
       "      <td>-0.147076</td>\n",
       "      <td>-0.226465</td>\n",
       "      <td>-0.013965</td>\n",
       "      <td>0.314038</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.325875</td>\n",
       "      <td>0.109699</td>\n",
       "      <td>-0.365506</td>\n",
       "      <td>0.646366</td>\n",
       "      <td>-0.360070</td>\n",
       "      <td>-0.081909</td>\n",
       "      <td>0.473251</td>\n",
       "      <td>-0.663408</td>\n",
       "      <td>0.628869</td>\n",
       "      <td>0.306305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.5</td>\n",
       "      <td>-0.723404</td>\n",
       "      <td>0.022613</td>\n",
       "      <td>-0.264118</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.791340</td>\n",
       "      <td>0.145261</td>\n",
       "      <td>0.105445</td>\n",
       "      <td>0.639288</td>\n",
       "      <td>1.300892</td>\n",
       "      <td>...</td>\n",
       "      <td>0.145265</td>\n",
       "      <td>1.123927</td>\n",
       "      <td>1.092905</td>\n",
       "      <td>-0.500491</td>\n",
       "      <td>-0.594256</td>\n",
       "      <td>-0.421169</td>\n",
       "      <td>0.091400</td>\n",
       "      <td>-0.547371</td>\n",
       "      <td>0.011724</td>\n",
       "      <td>0.607256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>350</th>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.531915</td>\n",
       "      <td>1.234051</td>\n",
       "      <td>0.124932</td>\n",
       "      <td>2.840187</td>\n",
       "      <td>0.594311</td>\n",
       "      <td>-0.155921</td>\n",
       "      <td>0.063501</td>\n",
       "      <td>10.464636</td>\n",
       "      <td>10.876975</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.035705</td>\n",
       "      <td>1.533840</td>\n",
       "      <td>-0.187597</td>\n",
       "      <td>-0.153438</td>\n",
       "      <td>0.215660</td>\n",
       "      <td>-1.240892</td>\n",
       "      <td>-0.280268</td>\n",
       "      <td>0.154121</td>\n",
       "      <td>0.502359</td>\n",
       "      <td>-0.038427</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>351</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.340426</td>\n",
       "      <td>3.358659</td>\n",
       "      <td>2.366177</td>\n",
       "      <td>0.709307</td>\n",
       "      <td>1.397244</td>\n",
       "      <td>0.219068</td>\n",
       "      <td>-0.012056</td>\n",
       "      <td>-0.058571</td>\n",
       "      <td>0.171578</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.095510</td>\n",
       "      <td>-0.878416</td>\n",
       "      <td>0.240232</td>\n",
       "      <td>-0.070995</td>\n",
       "      <td>0.064262</td>\n",
       "      <td>0.338604</td>\n",
       "      <td>-0.749398</td>\n",
       "      <td>0.023011</td>\n",
       "      <td>0.107523</td>\n",
       "      <td>0.435311</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>352</th>\n",
       "      <td>-0.5</td>\n",
       "      <td>0.468085</td>\n",
       "      <td>-0.008211</td>\n",
       "      <td>-0.313141</td>\n",
       "      <td>-0.413022</td>\n",
       "      <td>-0.283238</td>\n",
       "      <td>-0.184331</td>\n",
       "      <td>-0.232822</td>\n",
       "      <td>-0.210426</td>\n",
       "      <td>-0.219666</td>\n",
       "      <td>...</td>\n",
       "      <td>0.221995</td>\n",
       "      <td>0.064987</td>\n",
       "      <td>0.245384</td>\n",
       "      <td>-1.329974</td>\n",
       "      <td>-2.078832</td>\n",
       "      <td>-0.821568</td>\n",
       "      <td>-0.608044</td>\n",
       "      <td>-0.649634</td>\n",
       "      <td>-0.128411</td>\n",
       "      <td>4.613754</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>353</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.531915</td>\n",
       "      <td>1.188247</td>\n",
       "      <td>2.389373</td>\n",
       "      <td>0.340484</td>\n",
       "      <td>1.075466</td>\n",
       "      <td>1.630573</td>\n",
       "      <td>4.197880</td>\n",
       "      <td>8.086532</td>\n",
       "      <td>24.843820</td>\n",
       "      <td>...</td>\n",
       "      <td>0.062246</td>\n",
       "      <td>-0.158405</td>\n",
       "      <td>0.518511</td>\n",
       "      <td>0.206861</td>\n",
       "      <td>-0.119173</td>\n",
       "      <td>-0.592147</td>\n",
       "      <td>-0.076729</td>\n",
       "      <td>-0.796327</td>\n",
       "      <td>-0.379058</td>\n",
       "      <td>0.321494</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>354</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.382979</td>\n",
       "      <td>0.372430</td>\n",
       "      <td>0.047475</td>\n",
       "      <td>2.601662</td>\n",
       "      <td>7.822658</td>\n",
       "      <td>4.613486</td>\n",
       "      <td>1.496608</td>\n",
       "      <td>2.333308</td>\n",
       "      <td>0.244613</td>\n",
       "      <td>...</td>\n",
       "      <td>1.275178</td>\n",
       "      <td>-0.042744</td>\n",
       "      <td>-1.129216</td>\n",
       "      <td>-0.572312</td>\n",
       "      <td>0.509435</td>\n",
       "      <td>-0.219590</td>\n",
       "      <td>-0.161486</td>\n",
       "      <td>0.078554</td>\n",
       "      <td>0.100590</td>\n",
       "      <td>-0.127470</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>355 rows × 575 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Season  elapsed_time  E1A_mag_max_1  E1A_mag_max_2  E1A_mag_max_3  \\\n",
       "0       1.0      0.638298      -0.053077      -0.262788       0.047737   \n",
       "1      -0.5      0.680851      -0.072831      -0.266944      -0.349170   \n",
       "2       0.5     -0.446809      -0.075265      -0.297735      -0.293777   \n",
       "3       0.0     -0.510638       1.175205       5.244658       1.275429   \n",
       "4       0.5     -0.723404       0.022613      -0.264118       0.000000   \n",
       "..      ...           ...            ...            ...            ...   \n",
       "350     0.0     -0.531915       1.234051       0.124932       2.840187   \n",
       "351     0.5      0.340426       3.358659       2.366177       0.709307   \n",
       "352    -0.5      0.468085      -0.008211      -0.313141      -0.413022   \n",
       "353     0.0      0.531915       1.188247       2.389373       0.340484   \n",
       "354     1.0      0.382979       0.372430       0.047475       2.601662   \n",
       "\n",
       "     E1A_mag_max_4  E1A_mag_max_5  E1A_mag_max_6  E1A_mag_max_7  \\\n",
       "0        -0.242141      -0.169405      -0.221405      -0.215468   \n",
       "1        -0.152923       0.557963       8.095217      -0.112630   \n",
       "2        -0.248054      -0.185424      -0.242364      -0.142257   \n",
       "3        -0.107898      -0.147076      -0.226465      -0.013965   \n",
       "4         3.791340       0.145261       0.105445       0.639288   \n",
       "..             ...            ...            ...            ...   \n",
       "350       0.594311      -0.155921       0.063501      10.464636   \n",
       "351       1.397244       0.219068      -0.012056      -0.058571   \n",
       "352      -0.283238      -0.184331      -0.232822      -0.210426   \n",
       "353       1.075466       1.630573       4.197880       8.086532   \n",
       "354       7.822658       4.613486       1.496608       2.333308   \n",
       "\n",
       "     E1A_mag_max_8  ...  cor_E2P_M2P  cor_E3P_E4P  cor_E3P_M1P  cor_E3P_M2P  \\\n",
       "0        -0.232382  ...    -0.511122    -0.254089     0.096485    -0.548342   \n",
       "1         0.505699  ...     0.612548    -0.220519    -1.113509     0.697665   \n",
       "2        -0.231727  ...     0.368787    -1.128125     1.083483     0.035716   \n",
       "3         0.314038  ...    -0.325875     0.109699    -0.365506     0.646366   \n",
       "4         1.300892  ...     0.145265     1.123927     1.092905    -0.500491   \n",
       "..             ...  ...          ...          ...          ...          ...   \n",
       "350      10.876975  ...    -1.035705     1.533840    -0.187597    -0.153438   \n",
       "351       0.171578  ...    -0.095510    -0.878416     0.240232    -0.070995   \n",
       "352      -0.219666  ...     0.221995     0.064987     0.245384    -1.329974   \n",
       "353      24.843820  ...     0.062246    -0.158405     0.518511     0.206861   \n",
       "354       0.244613  ...     1.275178    -0.042744    -1.129216    -0.572312   \n",
       "\n",
       "     cor_E3V_E4A  cor_E3V_E4V  cor_E4A_E4H  cor_E4P_M2P  cor_M1A_M2A  \\\n",
       "0       1.769077     0.091947     0.726240    -0.476365     0.001820   \n",
       "1       2.088551     0.484775    -0.262615     0.257704    -0.460445   \n",
       "2      -1.100787     0.939328    -0.343666    -0.026715     1.450220   \n",
       "3      -0.360070    -0.081909     0.473251    -0.663408     0.628869   \n",
       "4      -0.594256    -0.421169     0.091400    -0.547371     0.011724   \n",
       "..           ...          ...          ...          ...          ...   \n",
       "350     0.215660    -1.240892    -0.280268     0.154121     0.502359   \n",
       "351     0.064262     0.338604    -0.749398     0.023011     0.107523   \n",
       "352    -2.078832    -0.821568    -0.608044    -0.649634    -0.128411   \n",
       "353    -0.119173    -0.592147    -0.076729    -0.796327    -0.379058   \n",
       "354     0.509435    -0.219590    -0.161486     0.078554     0.100590   \n",
       "\n",
       "     cor_M1P_M2P  \n",
       "0      -1.332654  \n",
       "1      -0.224991  \n",
       "2       1.111149  \n",
       "3       0.306305  \n",
       "4       0.607256  \n",
       "..           ...  \n",
       "350    -0.038427  \n",
       "351     0.435311  \n",
       "352     4.613754  \n",
       "353     0.321494  \n",
       "354    -0.127470  \n",
       "\n",
       "[355 rows x 575 columns]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Scale data with RobustScaler \n",
    "from sklearn.preprocessing import RobustScaler\n",
    "scaler = RobustScaler()\n",
    "col=train.columns\n",
    "X_tr_scaled = pd.DataFrame(scaler.fit_transform(train),columns=col)\n",
    "X_te_scaled = pd.DataFrame(scaler.fit_transform(test),columns=col )\n",
    "X_te_scaled"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dimensionality reduction with PCA (failed)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dimensionality Reduction with PCA (linear)\\\n",
    "\n",
    "In order to reduce our high dimensionality data with all the features extracted:\n",
    "- Peaks of magnitudes at LHC compressors low frequencies;\n",
    "- Autocorrelations among time series;\n",
    "- correlations one versus all.\n",
    "\n",
    "We tried to reduce the high dimensionality through a pincipal components analysis. PCA practically is a linear transformation of the variables that projects the original variables into a new Cartesian system in which the new variable with the highest variance is projected onto the first axis, the new variable, second in variance size, onto the second axis, and so on.\n",
    "We have tried to focus the reduction on the explained variance of the components, paying the price of lost information about specific columns (specific peaks of power spectrum for example.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Let's fit PCA\n",
    "pca = PCA().fit(X_tr_scaled)\n",
    "# display the percentage of total variance explained by each principal component#\n",
    "print(pca.explained_variance_ratio_.shape)"
   ]
  },
  {
   "attachments": {
    "num_comp.png": {
     "image/png": "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"
    },
    "scree_plot.png": {
     "image/png": "iVBORw0KGgoAAAANSUhEUgAAAm4AAAGJCAYAAAAzAb+0AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABJSElEQVR4nO3deVhU9f4H8PcwsiibKMoOg4qaC3pDQCzSriSWuZFLaqnkNTPXcCkrxaUuau7LT9tMb9etDM2y3AjN3PfU3EMBBXdAMEFnvr8/5jI5MuAZmOHMMO/X88yjnPOdM5/hXLvv57sdhRBCgIiIiIgsnp3cBRARERGRNAxuRERERFaCwY2IiIjISjC4EREREVkJBjciIiIiK8HgRkRERGQlGNyIiIiIrASDGxEREZGVYHAjIiIishIMbkREVmD58uVQKBS4dOmS3KUQkYwY3IjI6pw4cQI9evRAUFAQnJyc4OfnhxdeeAELFy6UuzSjTZ48GQqFQveqUaMGmjRpgg8//BB5eXkm+YxVq1Zh3rx5JrkWEcmrmtwFEBEZY8+ePXj++ecRGBiIwYMHw9vbGxkZGdi3bx/mz5+PESNGyF1iuSxZsgQuLi7Iz8/H1q1b8fHHH+OXX37B7t27oVAoKnTtVatW4eTJkxg9erRpiiUi2TC4EZFV+fjjj+Hu7o6DBw+iZs2aeueuX79e4esLIXD//n1Ur169wtcyRo8ePeDp6QkAeOutt/DKK68gOTkZ+/btQ1RUVKXWQkSWi0OlRGRVLl68iKZNm5YIbQBQt27dEsf++9//IiIiAjVq1ICHhweee+45bN26VXdepVLh5ZdfxpYtW9CqVStUr14dn376KQAgJycHo0ePRkBAABwdHdGgQQPMmDEDGo1G7zM0Gg3mzZuHpk2bwsnJCV5eXhgyZAju3LlT7u/5z3/+EwCQlpZWZrv/+7//Q9OmTeHo6AhfX18MGzYMOTk5uvPt2rXDpk2bcPnyZd1wrEqlKnddRCQv9rgRkVUJCgrC3r17cfLkSTRr1qzMtlOmTMHkyZPRpk0bTJ06FQ4ODti/fz9++eUXdOjQQdfu7Nmz6NOnD4YMGYLBgwejUaNGuHfvHtq2bYsrV65gyJAhCAwMxJ49ezBhwgRkZWXpzRkbMmQIli9fjvj4eIwcORJpaWlYtGgRjh49it27d8Pe3t7o73nx4kUAQO3atUttM3nyZEyZMgUxMTEYOnQozp49iyVLluDgwYO6z/3ggw+Qm5uLzMxMzJ07FwDg4uJidD1EZCEEEZEV2bp1q1AqlUKpVIqoqCgxfvx4sWXLFlFUVKTX7vz588LOzk50795dqNVqvXMajUb396CgIAFAbN68Wa/NtGnThLOzszh37pze8ffee08olUqRnp4uhBBi165dAoBYuXKlXrvNmzcbPP64xMREAUCcPXtW3LhxQ6SlpYlPP/1UODo6Ci8vL1FQUCCEEOKrr74SAERaWpoQQojr168LBwcH0aFDB73vt2jRIgFALFu2THesU6dOIigoqMw6iMg6cKiUiKzKCy+8gL1796JLly44fvw4Zs6cidjYWPj5+WHjxo26dhs2bIBGo8GkSZNgZ6f/n7rHJ/sHBwcjNjZW79i3336L6OhoeHh44ObNm7pXTEwM1Go1fv31V107d3d3vPDCC3rtwsLC4OLigtTUVEnfq1GjRqhTpw6Cg4MxZMgQNGjQAJs2bUKNGjUMtt++fTuKioowevRove83ePBguLm5YdOmTZI+l4isC4dKicjqhIeHIzk5GUVFRTh+/DjWr1+PuXPnokePHjh27BiaNGmCixcvws7ODk2aNHni9YKDg0scO3/+PH7//XfUqVPH4HuKF0KcP38eubm5BufXPdruSb777ju4ubnB3t4e/v7+qF+/fpntL1++DEAb+B7l4OCAevXq6c4TUdXC4EZEVsvBwQHh4eEIDw9Hw4YNER8fj2+//RaJiYlGXcfQClKNRoMXXngB48ePN/iehg0b6trVrVsXK1euNNiutOD3uOeee063qpSIqDQMbkRUJbRq1QoAkJWVBQCoX78+NBoN/vjjD7Rs2dLo69WvXx/5+fmIiYl5Yrvt27fjmWeeqdQtRIKCggBoF1bUq1dPd7yoqAhpaWl6dVd0Hzgishyc40ZEViU1NRVCiBLHf/rpJwB/Dx1269YNdnZ2mDp1aontOwy9/3G9evXC3r17sWXLlhLncnJy8PDhQ107tVqNadOmlWj38OFDva05TCkmJgYODg5YsGCB3vf58ssvkZubi06dOumOOTs7Izc31yx1EFHlYo8bEVmVESNG4N69e+jevTsaN26MoqIi7NmzB2vXroVKpUJ8fDwAoEGDBvjggw8wbdo0REdHIy4uDo6Ojjh48CB8fX2RlJRU5ueMGzcOGzduxMsvv4yBAwciLCwMBQUFOHHiBNatW4dLly7B09MTbdu2xZAhQ5CUlIRjx46hQ4cOsLe3x/nz5/Htt99i/vz56NGjh8l/D3Xq1MGECRMwZcoUdOzYEV26dMHZs2fxf//3fwgPD8drr72maxsWFoa1a9ciISEB4eHhcHFxQefOnU1eExFVAplXtRIRGeXnn38Wb7zxhmjcuLFwcXERDg4OokGDBmLEiBHi2rVrJdovW7ZM/OMf/xCOjo7Cw8NDtG3bVmzbtk13PigoSHTq1MngZ929e1dMmDBBNGjQQDg4OAhPT0/Rpk0bMWvWrBLbj3z22WciLCxMVK9eXbi6uormzZuL8ePHi6tXr5b5fYq3A7lx40aZ7R7fDqTYokWLROPGjYW9vb3w8vISQ4cOFXfu3NFrk5+fL/r27Stq1qwpAHBrECIrphBCwpgBEREREcmOc9yIiIiIrASDGxEREZGVYHAjIiIishIMbkRERERWgsGNiIiIyEowuBERERFZCW7Aa4BGo8HVq1fh6urKR8UQERGR2QkhcPfuXfj6+sLOrvR+NQY3A65evYqAgAC5yyAiIiIbk5GRAX9//1LPM7gZ4OrqCkD7y3Nzc5O5GiIiIqrq8vLyEBAQoMsgpWFwM6B4eNTNzY3BjYiIiCrNk6ZocXECERERkZVgcCMiIiKyEgxuRERERFaCwY2IiIjISjC4EREREVkJBjciIiIiK8HtQOSgVgO7dgFZWYCPDxAdDSiVcldFREREFo7BrbIlJwOjRgGZmX8f8/cH5s8H4uLkq4uIiIgsHodKK1NyMtCjh35oA4ArV7THk5PlqYuIiIisAoNbZVGrtT1tQpQ8V3xs9GhtOyIiIiIDGNwqy65dJXvaHiUEkJGhbUdERERkAINbZcnKMm07IiIisjkMbpXFx8e07YiIiMjmMLhVluho7epRhcLweYUCCAjQtiMiIiIygMGtsiiV2i0/gJLhrfjnefO4nxsRERGVisGtMsXFAevWAX5++sf9/bXHuY8bERERlYHBrbLFxQGXLgHffKP9eetWIC2NoY2IiIieiMFNDkol8MorgIMDoFJxeJSIiIgkYXCTi50dEBwM/Pmn3JUQERGRlWBwk1O9egxuREREJBmDm5wY3IiIiMgIDG5yYnAjIiIiIzC4yYnBjYiIiIzA4Can4sUJQshdCREREVkBBjc5BQcDeXnA7dtyV0JERERWgMFNTm5ugKcnh0uJiIhIEgY3uXGeGxEREUnE4CY3BjciIiKSiMFNbgxuREREJBGDm9zq1dM+ZJ6IiIjoCRjc5MYeNyIiIpKIwU1u9eoB6enAgwdyV0JEREQWjsFNbv7+gEIBZGTIXQkRERFZOAY3uSmVgErF4VIiIiJ6IgY3S8B5bkRERCQBg5slYHAjIiIiCRjcLAGDGxEREUnA4GYJGNyIiIhIAgY3S8DgRkRERBIwuFmCevWAO3e0LyIiIqJSMLhZAnd3oFYtPvqKiIiIysTgZik4XEpERERPwOBmKRjciIiI6AkY3CxFcDCDGxEREZWJwc1SsMeNiIiInoDBzVIwuBEREdETMLhZinr1gMuXgYcP5a6EiIiILBSDm6UICACEADIz5a6EiIiILBSDm6WwtwcCAzlcSkRERKVicLMknOdGREREZWBwsyQMbkRERFQGiwhuixcvhkqlgpOTEyIjI3HgwIFS237++eeIjo6Gh4cHPDw8EBMTU6K9EAKTJk2Cj48PqlevjpiYGJw/f97cX6PiGNyIiIioDLIHt7Vr1yIhIQGJiYk4cuQIWrRogdjYWFy/ft1g+x07dqBPnz5ITU3F3r17ERAQgA4dOuDKlSu6NjNnzsSCBQuwdOlS7N+/H87OzoiNjcX9+/cr62uVD4MbERERlUEhhBByFhAZGYnw8HAsWrQIAKDRaBAQEIARI0bgvffee+L71Wo1PDw8sGjRIvTv3x9CCPj6+mLMmDEYO3YsACA3NxdeXl5Yvnw5Xn311SdeMy8vD+7u7sjNzYWbm1vFvqAxDh0COnYEbt6svM8kIiIi2UnNHrL2uBUVFeHw4cOIiYnRHbOzs0NMTAz27t0r6Rr37t3DgwcPUKtWLQBAWloasrOz9a7p7u6OyMjIUq9ZWFiIvLw8vZcs6tUDbt0C5Pp8IiIismiyBrebN29CrVbDy8tL77iXlxeys7MlXePdd9+Fr6+vLqgVv8+YayYlJcHd3V33CggIMParmIaHB+DuDqSlyfP5REREZNFkn+NWEdOnT8eaNWuwfv16ODk5lfs6EyZMQG5uru6VkZFhwiqNoFBwnhsRERGVqpqcH+7p6QmlUolr167pHb927Rq8vb3LfO+sWbMwffp0bN++HaGhobrjxe+7du0afHx89K7ZsmVLg9dydHSEo6NjOb+FiTG4ERERUSlk7XFzcHBAWFgYUlJSdMc0Gg1SUlIQFRVV6vtmzpyJadOmYfPmzWjVqpXeueDgYHh7e+tdMy8vD/v37y/zmhaDwY2IiIhKIWuPGwAkJCRgwIABaNWqFSIiIjBv3jwUFBQgPj4eANC/f3/4+fkhKSkJADBjxgxMmjQJq1atgkql0s1bc3FxgYuLCxQKBUaPHo2PPvoIISEhCA4OxsSJE+Hr64tu3brJ9TWlq1cP+P57uasgIiIiCyR7cOvduzdu3LiBSZMmITs7Gy1btsTmzZt1iwvS09NhZ/d3x+CSJUtQVFSEHj166F0nMTERkydPBgCMHz8eBQUFePPNN5GTk4Nnn30WmzdvrtA8uErDHjciIiIqhez7uFki2fZxA4ALF4CmTYF79wClsnI/m4iIiGRhFfu4kQGBgcDDh8DVq3JXQkRERBaGwc3SODgAAQEcLiUiIqISGNwsUXAwgxsRERGVwOBmibhAgYiIiAxgcLNEDG5ERERkAIObJWJwIyIiIgMY3CwRgxsREREZwOBmierVA65fB/Lz5a6EiIiILAiDmyXy9ARcXIC0NLkrISIiIgvC4GaJFAoOlxIREVEJDG6WisGNiIiIHsPgZqkY3IiIiOgxDG6WisGNiIiIHsPgZqkY3IiIiOgxDG6Wql497apSjUbuSoiIiMhCMLhZqqAgoKgIyMqSuxIiIiKyEAxulsrJCfDz43ApERER6TC4WTLOcyMiIqJHMLhZMgY3IiIiegSDmyVjcCMiIqJHMLhZMpUKOHYMWL0a2LEDUKtlLoiIiIjkxOBmqZKTgbFjgZMngb59geef1wa55GS5KyMiIiKZMLhZouRkoEcP4Pp1/eNXrmiPM7wRERHZJAY3S6NWA6NGAUKUPFd8bPRoDpsSERHZIAY3S7NrF5CZWfp5IYCMDG07IiIisinVpDRKSEiQfME5c+aUuxiC9Ccl8IkKRERENkdScDt69Kjez0eOHMHDhw/RqFEjAMC5c+egVCoRFhZm+gptjY+PadsRERFRlSEpuKWmpur+PmfOHLi6umLFihXw8PAAANy5cwfx8fGIjo42T5W2JDoa8PfXLkQwNM9NodCe5++aiIjI5iiEMJQOSufn54etW7eiadOmesdPnjyJDh064OrVqyYtUA55eXlwd3dHbm4u3NzcKr+A4lWlgH54Uyi0f65bB8TFVX5dREREZBZSs4fRixPy8vJw48aNEsdv3LiBu3fvGns5MiQuThvO/Pz0j/v7M7QRERHZMKODW/fu3REfH4/k5GRkZmYiMzMT3333HQYNGoQ4BgrTiYsDLl0CNm7U/vzDD0BaGkMbERGRDZM0x+1RS5cuxdixY9G3b188ePBAe5Fq1TBo0CB88sknJi/QpimVQOfOgLu7trdNqZS7IiIiIpKR0XPcihUUFODixYsAgPr168PZ2dmkhclJ9jluj2vZEkhMBLp3l7sSIiIiMgOzzXErlpWVhaysLISEhMDZ2RnlzH8kRXCwdtiUiIiIbJrRwe3WrVto3749GjZsiJdeeglZ/9sIdtCgQRgzZozJCyRoHy6fliZ3FURERCQzo4PbO++8A3t7e6Snp6NGjRq6471798bmzZtNWhz9D3vciIiICOVYnLB161Zs2bIF/v7+esdDQkJw+fJlkxVGj2CPGxEREaEcPW4FBQV6PW3Fbt++DUdHR5MURY8p7nHjPEIiIiKbZnRwi46Oxn/+8x/dzwqFAhqNBjNnzsTzzz9v0uLof1QqID8fuHVL7kqIiIhIRkYPlc6cORPt27fHoUOHUFRUhPHjx+PUqVO4ffs2du/ebY4aydUVqF1b2+vm6Sl3NURERCQTo3vcmjVrhnPnzuHZZ59F165dUVBQgLi4OBw9ehT169c3R40EaIdLOc+NiIjIphnd4wYA7u7u+OCDD0xdC5VFpeLKUiIiIhtXruCWk5ODAwcO4Pr169BoNHrn+vfvb5LC6DHscSMiIrJ5Rge3H374Af369UN+fj7c3NygUCh05xQKBYObuahUwI8/yl0FERERycjoOW5jxozBG2+8gfz8fOTk5ODOnTu61+3bt81RIwHscSMiIiLjg9uVK1cwcuRIg3u5kRkVz3HjXm5EREQ2y+jgFhsbi0OHDpmjFiqLSgXcvw9cuyZ3JURERCQTo+e4derUCePGjcMff/yB5s2bw97eXu98ly5dTFYcPaJ6dcDLS9vr5u0tdzVEREQkA4UQxo292dmV3kmnUCigVqsrXJTc8vLy4O7ujtzcXLi5ucldzt+iooCRI4E+feSuhIiIiExIavYweqhUo9GU+ipPaFu8eDFUKhWcnJwQGRmJAwcOlNr21KlTeOWVV6BSqaBQKDBv3rwSbSZPngyFQqH3aty4sdF1WSTu5UZERGTTjA5uprR27VokJCQgMTERR44cQYsWLRAbG4vr168bbH/v3j3Uq1cP06dPh3cZw4VNmzZFVlaW7vXbb7+Z6ytULq4sJSIismmS5rgtWLAAb775JpycnLBgwYIy244cOVLyh8+ZMweDBw9GfHw8AGDp0qXYtGkTli1bhvfee69E+/DwcISHhwOAwfPFqlWrVmaws1oqFcCFIURERDZLUnCbO3cu+vXrBycnJ8ydO7fUdgqFQnJwKyoqwuHDhzFhwgTdMTs7O8TExGDv3r2SrlGa8+fPw9fXF05OToiKikJSUhICAwNLbV9YWIjCwkLdz3l5eRX6fLNhjxsREZFNkxTc0h4JC2kmCg43b96EWq2Gl5eX3nEvLy+cOXOm3NeNjIzE8uXL0ahRI2RlZWHKlCmIjo7GyZMn4erqavA9SUlJmDJlSrk/s9KoVMDly4BGA5SxSISIiIiqpir3//4vvvgievbsidDQUMTGxuKnn35CTk4Ovvnmm1LfM2HCBOTm5upeGRkZlVixEQIDgYcPgatX5a6EiIiIZFCuh8xnZmZi48aNSE9PR1FRkd65OXPmSLqGp6cnlEolrj22oey1a9dMOj+tZs2aaNiwIS5cuFBqG0dHRzg6OprsM83G0RHw9dWuLPX3l7saIiIiqmRGB7eUlBR06dIF9erVw5kzZ9CsWTNcunQJQgg8/fTTkq/j4OCAsLAwpKSkoFu3bgC0W42kpKRg+PDhxpZVqvz8fFy8eBGvv/66ya4pq+J5bs8+K3clREREVMmMHiqdMGECxo4dixMnTsDJyQnfffcdMjIy0LZtW/Ts2dOoayUkJODzzz/HihUrcPr0aQwdOhQFBQW6Vab9+/fXW7xQVFSEY8eO4dixYygqKsKVK1dw7Ngxvd60sWPHYufOnbh06RL27NmD7t27Q6lUok9V2bSWe7kRERHZLKN73E6fPo3Vq1dr31ytGv766y+4uLhg6tSp6Nq1K4YOHSr5Wr1798aNGzcwadIkZGdno2XLlti8ebNuwUJ6errekxquXr2Kf/zjH7qfZ82ahVmzZqFt27bYsWMHAO0wbp8+fXDr1i3UqVMHzz77LPbt24c6deoY+1UtE1eWEhER2Syjg5uzs7NuXpuPjw8uXryIpk2bAtCuFDXW8OHDSx0aLQ5jxVQqFZ70hK41a9YYXYNVUamAqrKhMBERERnF6ODWunVr/Pbbb3jqqafw0ksvYcyYMThx4gSSk5PRunVrc9RIj2KPGxERkc0yOrjNmTMH+fn5AIApU6YgPz8fa9euRUhIiOQVpVQBKhWQkaHdFqRauRYFExERkZVSiCeNPdqgvLw8uLu7Izc3F25ubnKXo+/hQ8DJCbhwQRviiIiIyOpJzR5VbgPeKq9aNe0eblxZSkREZHMkjbV5eHhAoVBIuuDt27crVBBJUDzPrV07uSshIiKiSiQpuM2bN8/MZZBRuJcbERGRTZIU3AYMGGDuOsgYwcHaOW5ERERkU8q1LFGtVmP9+vU4ffo0AKBJkybo2rUrqnGVY+VQqYDt2+WugoiIiCqZ0Unr1KlT6NKlC7Kzs9GoUSMAwIwZM1CnTh388MMPaNasmcmLpMdwLzciIiKbZPSq0n/9619o2rQpMjMzceTIERw5cgQZGRkIDQ3Fm2++aY4a6XEqFXDlCvC/J1gQERGRbTC6x+3YsWM4dOgQPDw8dMc8PDzw8ccfIzw83KTFUSl8fbXbgqSnAw0ayF0NERERVRKje9waNmyIa9eulTh+/fp1NGCIqBxKJRAYyJWlRERENsbo4JaUlISRI0di3bp1yMzMRGZmJtatW4fRo0djxowZyMvL073IjDjPjYiIyOYYPVT68ssvAwB69eql25S3+KlZnTt31v2sUCigVqtNVSc9jnu5ERER2Ryjg1tqaqo56iBjBQcDJ0/KXQURERFVIqODW9u2bc1RBxlLpQJ+/FHuKoiIiKgSGT3HbfLkydBoNCWO5+bmok+fPiYpiiTgHDciIiKbY3Rw+/LLL/Hss8/izz//1B3bsWMHmjdvjosXL5q0OCqDSgVkZwN//SV3JURERFRJjA5uv//+O/z9/dGyZUt8/vnnGDduHDp06IDXX38de/bsMUeNZIi3N+DkBFy+LHclREREVEmMnuPm4eGBb775Bu+//z6GDBmCatWq4eeff0b79u3NUR+VRqEAgoK0K0sbN5a7GiIiIqoERve4AcDChQsxf/589OnTB/Xq1cPIkSNx/PhxU9dGT8J5bkRERDbF6ODWsWNHTJkyBStWrMDKlStx9OhRPPfcc2jdujVmzpxpjhqpNNzLjYiIyKYYHdzUajV+//139OjRAwBQvXp1LFmyBOvWrcPcuXNNXiCVgT1uRERENsXo4LZt2zb4+vqWON6pUyecOHHCJEWRROxxIyIisimSg9uBAwfKfIRVYWEhfvnlF5MURRKxx42IiMimSA5uUVFRuHXrlu5nNzc3vb3ccnJyuAFvZVOpgJs3gfx8uSshIiKiSiA5uBU/SL60n0s7Rmbk6Qk4O3O4lIiIyEaUazuQ0igUClNejp5EoeA8NyIiIhti0uBGMuA8NyIiIpth1JMT/vjjD2RnZwPQDoueOXMG+f+bX3Xz5k3TV0dPxh43IiIim2FUcGvfvr3ePLaXX34ZgHaIVAjBoVI5BAcDv/0mdxVERERUCSQHtzQOx1kmlQr473/lroKIiIgqgeTgFhQUZM46qLw4x42IiMhmcHGCtVOpgJwc7YuIiIiqNAY3a+fhAbi7c4ECERGRDWBwqwq4spSIiMgmMLhVBZznRkREZBPKFdwePnyI7du349NPP8Xdu3cBAFevXtXt6UaVjD1uRERENsGofdwA4PLly+jYsSPS09NRWFiIF154Aa6urpgxYwYKCwuxdOlSc9RJZQkOBrZvl7sKIiIiMjOje9xGjRqFVq1a4c6dO6hevbruePfu3ZGSkmLS4kiigADg5Elg9Wpgxw5ArZa7IiIiIjIDo3vcdu3ahT179sDBwUHvuEqlwpUrV0xWGEmUnAy8/TZw7RrQt6/2mL8/MH8+EBcnb21ERERkUkb3uGk0GqgN9OhkZmbC1dXVJEWRRMnJQI8e2tD2qCtXtMeTk+Wpi4iIiMzC6ODWoUMHzJs3T/ezQqFAfn4+EhMT8dJLL5myNiqLWg2MGgU88uxYneJjo0dz2JSIiKgKMTq4zZ49G7t370aTJk1w//599O3bVzdMOmPGDHPUSIbs2gVkZpZ+XgggI0PbjoiIiKoEo+e4+fv74/jx41i7di2OHz+O/Px8DBo0CP369dNbrEBmlpVl2nZERERk8YwObgBQrVo19OvXD/369TN1PSSVj49p2xEREZHFM3qoNCkpCcuWLStxfNmyZRwqrUzR0drVowqF4fMKhXabkOjoyq2LiIiIzMbo4Pbpp5+icePGJY43bdqUm+9WJqVSu+UHUDK8Ff88b562HREREVUJRge37Oxs+BgYfqtTpw6yyjGfavHixVCpVHByckJkZCQOHDhQattTp07hlVdegUqlgkKh0FvdWt5rWrW4OGDdOsDPT/+4v7/2OPdxIyIiqlKMDm4BAQHYvXt3ieO7d++Gr6+vUddau3YtEhISkJiYiCNHjqBFixaIjY3F9evXDba/d+8e6tWrh+nTp8Pb29sk17R6cXHa55SmpgKNGgHjx2sfOM/QRkREVOUYHdwGDx6M0aNH46uvvsLly5dx+fJlLFu2DO+88w4GDx5s1LXmzJmDwYMHIz4+Hk2aNMHSpUtRo0YNg3PoACA8PByffPIJXn31VTg6OprkmlWCUgm0awfExAAPHnB4lIiIqIoyelXpuHHjcOvWLbz99tsoKioCADg5OeHdd9/FhAkTJF+nqKgIhw8f1nuPnZ0dYmJisHfvXmPLqtA1CwsLUVhYqPs5Ly+vXJ8vu+bNtUOkREREVCUZ3eOmUCgwY8YM3LhxA/v27cPx48dx+/ZtTJo0yajr3Lx5E2q1Gl5eXnrHvby8kJ2dbWxZFbpmUlIS3N3dda+AgIByfb7smjcHTpyQuwoiIiIyE6ODWzEXFxeEh4ejWbNmpQ5bWosJEyYgNzdX98rIyJC7pPJp1kz73NKqOp+PiIjIxhk9VFpQUIDp06cjJSUF169fh0aj0Tv/559/SrqOp6cnlEolrj32gPRr166VuvDAXNd0dHS0+vAJAHBzA1Qqba9b+/ZyV0NEREQmZnRw+9e//oWdO3fi9ddfh4+PDxSlbQD7BA4ODggLC0NKSgq6desGANBoNEhJScHw4cMt5ppWp3i4lMGNiIioyjE6uP3888/YtGkTnnnmmQp/eEJCAgYMGIBWrVohIiIC8+bNQ0FBAeLj4wEA/fv3h5+fH5KSkgBoFx/88ccfur9fuXIFx44dg4uLCxo0aCDpmlUe57kRERFVWUYHNw8PD9SqVcskH967d2/cuHEDkyZNQnZ2Nlq2bInNmzfrFhekp6fDzu7vaXhXr17FP/7xD93Ps2bNwqxZs9C2bVvs2LFD0jWrvNBQYOtWuasgIiIiM1AIIYQxb/jvf/+L77//HitWrECNGjXMVZes8vLy4O7ujtzcXLi5ucldjnH++ANo1Qq4e5f7uREREVkJqdnD6B632bNn4+LFi/Dy8oJKpYK9vb3e+SNHjhhfLZlOSAigVgN//qn9OxEREVUZRge34kn/ZKHs7YGnngJ+/53BjYiIqIoxOrglJiaaow4ypdBQ7QKFV16RuxIiIiIyoXJvwEsWjCtLiYiIqiSjg5tarcasWbMQEREBb29v1KpVS+9FFoDBjYiIqEoyOrhNmTIFc+bMQe/evZGbm4uEhATExcXBzs4OkydPNkOJZLTQUODCBaCgQO5KiIiIyISMDm4rV67E559/jjFjxqBatWro06cPvvjiC0yaNAn79u0zR41kLB8fwMNDuzUIERERVRlGB7fs7Gw0b94cgPZB87m5uQCAl19+GZs2bTJtdVQ+CgWHS4mIiKogo4Obv78/srKyAAD169fH1v/t0n/w4MGq8aD2qoLBjYiIqMoxOrh1794dKSkpAIARI0Zg4sSJCAkJQf/+/fHGG2+YvEAqp9BQ7V5uREREVGUY/cirx+3duxd79+5FSEgIOnfubKq6ZGXVj7wqtm8f0KULcP263JUQERHRE0jNHhUOblVRlQhud+8Cbm5Adjbg5SV3NURERFQGkz6rdOPGjXjxxRdhb2+PjRs3ltm2S5cuxlVK5uHqCgQHa+e5MbgRERFVCZKCW7du3ZCdnY26deuW+axShUIBtVptqtqooornucXEyF0JERERmYCk4KbRaAz+nSwcV5YSERFVKUatKn3w4AHat2+P8+fPm6seMiUGNyIioirFqOBmb2+P37nFhPUIDQVOnQI4fE1ERFQlGL2P22uvvYYvv/zSHLWQqTVoAAihfW4pERERWT1Jc9we9fDhQyxbtgzbt29HWFgYnJ2d9c7PmTPHZMVRBVWrBjRpoh0ubdRI7mqIiIiogowObidPnsTTTz8NADh37pzeOYVCYZqqyHSK57n16CF3JURERFRBRge31NRUc9RB5hIaCvz2m9xVEBERkQkYPceNrAxXlhIREVUZRve4AcChQ4fwzTffID09HUVFRXrnkpOTTVIYmUjz5sCffwIFBcBj8xGJiIjIuhjd47ZmzRq0adMGp0+fxvr16/HgwQOcOnUKv/zyC9zd3c1RI1WEtzdQu7Z2WxAiIiKyakYHt3//+9+YO3cufvjhBzg4OGD+/Pk4c+YMevXqhcDAQHPUSBWhUGh73bj/HhERkdUzOrhdvHgRnTp1AgA4ODigoKAACoUC77zzDj777DOTF0gmwHluREREVYLRwc3DwwN3794FAPj5+eHkyZMAgJycHNy7d8+01ZFpMLgRERFVCUYHt+eeew7btm0DAPTs2ROjRo3C4MGD0adPH7Rv397kBZIJhIZqh0qFkLsSIiIiqgCFENL+3/zkyZNo1qwZbt++jfv378PX1xcajQYzZ87Enj17EBISgg8//BAeHh7mrtns8vLy4O7ujtzcXLi5ucldTsUVFACursCVK4CPj9zVEBER0WOkZg/J24GEhoYiPDwc//rXv/Dqq68CAOzs7PDee+9VvFoyL2dnoF497XApgxsREZHVkjxUunPnTjRt2hRjxoyBj48PBgwYgF27dpmzNjKl0FDOcyMiIrJykoNbdHQ0li1bhqysLCxcuBCXLl1C27Zt0bBhQ8yYMQPZ2dnmrJMqiluCEBERWT2jFyc4OzsjPj4eO3fuxLlz59CzZ08sXrwYgYGB6NKlizlqJFPgylIiIiKrJ3lxQmkKCgqwcuVKTJgwATk5OVCr1aaqTTZVbnECAJw9C7RoAeTnA9XK9aQzIiIiMhOp2aPcD5n/9ddfMXDgQHh7e2PcuHGIi4vD7t27y3s5MrcGDbRPUbhwQe5KiIiIqJyM6nq5evUqli9fjuXLl+PChQto06YNFixYgF69esGZDzC3bEol0LSpdp5b48ZyV0NERETlIDm4vfjii9i+fTs8PT3Rv39/vPHGG2jUqJE5ayNTK57n1quX3JUQERFROUgObvb29li3bh1efvllKJVKc9ZE5tKsGbB+PbB6tXY/t+hobU8cERERWYUKL06oiqrk4oTkZGDIEODmzb+P+fsD8+cDcXHy1UVERETmX5xAViQ5GejRQz+0AdpHYPXooT1PREREFo/BrapTq4FRoww/YL742OjR2nZERERk0Rjcqrpdu4DMzNLPCwFkZGjbERERkUVjcKvqsrJM246IiIhkw+BW1fn4mLYdERERyYbBraqLjtauHlUoDJ9XKICAAG07IiIismgMblWdUqnd8gMoPbzNm8f93IiIiKwAg5stiIsD1q0D/PxKnpszh/u4ERERWQluwGtAldyAF9Bu+bFrl3Yhgo+Pdv+28+eBn3+WuzIiIiKbZlUb8C5evBgqlQpOTk6IjIzEgQMHymz/7bffonHjxnByckLz5s3x008/6Z0fOHAgFAqF3qtjx47m/ArWQakE2rUD+vTR/jl5MrBvH7B1q8yFERERkRSyB7e1a9ciISEBiYmJOHLkCFq0aIHY2Fhcv37dYPs9e/agT58+GDRoEI4ePYpu3bqhW7duOHnypF67jh07IisrS/davXp1ZXwd61KrFjBxIjBmDDfgJSIisgKyD5VGRkYiPDwcixYtAgBoNBoEBARgxIgReO+990q07927NwoKCvDjjz/qjrVu3RotW7bE0qVLAWh73HJycrBhw4Zy1VRlh0oNKSwEmjYF3n0XGDxY7mqIiIhsklUMlRYVFeHw4cOIiYnRHbOzs0NMTAz27t1r8D179+7Vaw8AsbGxJdrv2LEDdevWRaNGjTB06FDcunWr1DoKCwuRl5en97IZjo7AjBnanre7d+WuhoiIiMoga3C7efMm1Go1vLy89I57eXkhOzvb4Huys7Of2L5jx474z3/+g5SUFMyYMQM7d+7Eiy++CHUpw4FJSUlwd3fXvQICAir4zaxMXBwQEgLMnCl3JURERFSGanIXYA6vvvqq7u/NmzdHaGgo6tevjx07dqB9+/Yl2k+YMAEJCQm6n/Py8mwrvCkUwOzZwPPPA2++qd2Ql4iIiCyOrD1unp6eUCqVuHbtmt7xa9euwdvb2+B7vL29jWoPAPXq1YOnpycuXLhg8LyjoyPc3Nz0XjYnIgLo1g344AO5KyEiIqJSyBrcHBwcEBYWhpSUFN0xjUaDlJQUREVFGXxPVFSUXnsA2LZtW6ntASAzMxO3bt2CD5/HWbZ//1u7Ue+BA8COHcDq1do/ueKUiIjIIsg+VJqQkIABAwagVatWiIiIwLx581BQUID4+HgAQP/+/eHn54ekpCQAwKhRo9C2bVvMnj0bnTp1wpo1a3Do0CF89tlnAID8/HxMmTIFr7zyCry9vXHx4kWMHz8eDRo0QGxsrGzf0yoEBQEdO2qfW1pU9Pdxf3/tY7P4hAUiIiJZyR7cevfujRs3bmDSpEnIzs5Gy5YtsXnzZt0ChPT0dNjZ/d0x2KZNG6xatQoffvgh3n//fYSEhGDDhg1o1qwZAECpVOL333/HihUrkJOTA19fX3To0AHTpk2Do6OjLN/RaiQnAxs2AI/vEHPlCtCjh7Y3juGNiIhINrLv42aJbGoft2JqNaBSAZmZhs8rFNqet7Q0PpCeiIjIxKxiHzeyILt2lR7aAG0vXEaGtl0xtZpz4YiIiCqR7EOlZCGysoxrl5wMjBqlH/Y4F46IiMisGNxIS+qK299+Ax4+BAYM4Fw4IiKiSsahUtKKjtb2mCkUhs8rFEDt2sCRI4ZDG/D3sdGjOWxKRERkBgxupKVUaoc5gZLhrfjnzz4DkpIMh7ZihubCERERkUkwuNHf4uK0w5x+fvrH/f3/Hv40di4cERERmQznuJG+uDiga1dtj1lWlnbuW3T031uASJ0Lx6dUEBERmRyDG5WkVALt2hk+VzwX7soVw0Omxfu9RUebtUQiIiJbxKFSMk5Zc+GKzZvHTXqJiIjMgMGNjFfaXDilEhg8mFuBEBERmQkfeWWATT7yqjzUav25cPn5wGuvAWfOAN7ecldHRERkNaRmD85xo/IzNBeuY0fgnXe0j8EiIiIik+JQKZnW3LnAzz8DW7bIXQkREVGVw+BGpuXjo92kd+hQ4N49uashIiKqUhjcyPSGDAHq1gU++kjuSoiIiKoUBjcyPTs77eOx5s0DTp6UuxoiIqIqg8GNzCM0FBg+HHjrLeDBA2DHDu2ChR07+AB6IiKicuKqUjKfxEQgOBjw8gLu3Pn7uL+/dhNf7vdGRERkFPa4kfls2QLcuKEf2gDt47J69ACSk+Wpi4iIyEoxuJF5qNXAqFGGzxXv+Tx6NIdNiYiIjMDgRuaxaxeQmVn6eSGAjAxtu2JqNefCERERlYFz3Mg8srKMa5ecrO2hezTslTYX7vFHbUVH86H2RERkExjcyDx8fKS1O3gQKCoC4uP/HkItVjwXbt26v8ObMQGPiIioiuFD5g3gQ+ZNQK0GVCpt+DL0PzGFAnB3Bxo2BA4cKP06CoU2mKWlAd9/rw1yj19PodD++WjAIyIisiJSswfnuJF5KJXaXjDg72BVrPjnL78EZswo+zrFc+EWLgSGDTMcArnYgYiIbASDG5lPXJy2F8zPT/+4v//fvWNS58IlJgLZ2aWfN7TYgYiIqIrhHDcyr7g4oGvX0hcTSJ0LN3KktGefPhoEuYiBiIiqGAY3Mj+lEmjXzvC56GhtD1xZc+H8/bXvlxLciq/BRQxERFQFcaiU5CVlLty8edrg5u9fss2jbWvU0K5O7dhRu4jh8X3k+MQGIiKycgxuJD8pc+GkBLyvvwYOHQJ+/dW4RQzc+JeIiKwEgxtZhrg44NIlIDUVWLVK+2damv6wppSAd+sW8NdfpX/O44sYkpO125Y8/zzQt6/2T5WKvXJERGSRuI+bAdzHzcKVtehg9WptAHuS//wHcHY2bl84LnYgIiIzkZo9GNwMYHCzYjt2aHvNnqRGDW1gK6137tGNf5VKLnYgIiKz4ga8ZJuKV6mWtYghIAD45BPpQ6rJycYtduCcOSIiMhMGN6papK5S9fCQdr0FC4A335S+2MHYOXNSQx7DIBERgcGNqiIpixikbvybmald8FCaivTMSQ15xoRBBjwioiqNc9wM4By3KqKsxQRqtTb8PGnj36Qk4LXXnvxZ//wncPQocOeO4fOG5sxJWRghtR3AeXhERFaMixMqgMHNRhSHIkA/GD0aimrVkrbYoV07bQ/XkyxdCrzwAvDcc9rQaEhxyLtwAahfv2QP3uPt0tKA77/nClkiIivG4FYBDG42xFAvVUCAdh5cXJzpe+ZcXYG7d6XVFhkJ7N//5HbbtwMDB0oLeMaukJUa8BgEiYgqhMGtAhjcbMyTQocpe+ZSU7UhUErIa9VK+ySIJwkMBNLTpX327dumH3o1RxAkIrIx3A6ESCqlUjvU2aeP9s/Hg4SUxQ5StyGJji55ndL06yetXYsW0tpdvKgNWFJWyEpdaGHMggwusiAiqjD2uBnAHjcyyBQ9c8YMvxbPcXtSu6++AmJiTPY1JQ+9mmsOHodyicgGSc4egkrIzc0VAERubq7cpZC1+e47Ifz9hdBGFO0rIEB7/PF2CoX29Wjb4mPF7aW0e/hQ+5mPt3m0bUCAEPPnGz7/+MvbW1q7t96S1m779pK/E0P1PXz49/c11ObR30tpv2t/f8O/aynthNDWkJoqxKpV2j8fPjR8n+VqR0RVltTsweBmAIMbVYjU/xM2JuQ9qZ2UgJeaKi1o9eghrV3NmtLatWghX8CTMwjKGSyNactwSWQRGNwqgMGNKo0p/0/zSQFPas/c9u3SgtbcudLaPfOMtHYhIaYNeIWF8gVBOYOlMW1tsdeSgZYsFINbBTC4kdV60v+JmHLotTgYmSoIxsZKayd1KHfYMHmCoKnblWcI2dLDpa21k/Jvs6q1M9c1qzAGtwpgcKMqzVRDr1Lbmbqn75VXpLVzdZXWztFRWruICGntevWS1m7kSGntNmwQwtdXWsAr/l0/qa2c4dLW2pX2b86SgqU5gqotht8KYnCrAAY3qvJMMfRqTDs5Ap7UodzBg6W1Cw+X1i4gQFo7Z2dp7aS+6tcXokkTaW2ff15au27dpLVbuVJauJSr19IaekurSjtzXtPSw2oFMbhVAIMb0f9U5hy84jaWOpQrNQiaut24cdLajRkjvReveXNp7YKDpbWT+qpdW1q7Ro2ktZMaQKX2gkoN8e++K63dDz8I4edX+nm5g6U5gqqpe32tIfyaiFUFt0WLFomgoCDh6OgoIiIixP79+8ts/80334hGjRoJR0dH0axZM7Fp0ya98xqNRkycOFF4e3sLJycn0b59e3Hu3DnJ9TC4EZlJZQY8qe1MHQRN3U5qsExNlb5y2NTh8oMPpLWT2oPXrp20ds2aSWsntRfUw0Nau2rVpLWT+nJyMu33kBrMn35aWrvnnpPWbtQoIRISpLWVup3Q2rWW3ZtrwmFTqwlua9asEQ4ODmLZsmXi1KlTYvDgwaJmzZri2rVrBtvv3r1bKJVKMXPmTPHHH3+IDz/8UNjb24sTJ07o2kyfPl24u7uLDRs2iOPHj4suXbqI4OBg8ddff0mqicGNSGaWOJQrVzupwfLR3g5LDZdy9VrK1W7MGGnteveW1k7qAp5nn5XWLipKWruWLaW1+8c/hAgNldbW3V1aO6kvOztp7VxcpLXz9JTWLjXVVP/Vs57gFhERIYYNG6b7Wa1WC19fX5GUlGSwfa9evUSnTp30jkVGRoohQ4YIIbS9bd7e3uKTTz7Rnc/JyRGOjo5i9erVBq95//59kZubq3tlZGRI+uURkcwqeyhXrnZSg6AxbW2p19LSe0stPYBKbWeOXt/335fW7vXXpbWTuk9lly7S2q1aZfi/OeVgFcGtsLBQKJVKsX79er3j/fv3F126dDH4noCAADF37ly9Y5MmTRKhoaFCCCEuXrwoAIijR4/qtXnuuefEyJEjDV4zMTFRACjxYnAjsjGWvGpNahA0pq2t9FrK1c7WAq05en0tPfzaWo/blStXBACxZ88evePjxo0TERERBt9jb28vVj2WcBcvXizq1q0rhNAOpQIQV69e1WvTs2dP0atXL4PXZI8bEVkFS39ygiX3WsrVzpKDpTnamfqalh5+bW2Om6UEt8dxjhsRUTlZcq+lXO0sOViao52pr2kNYdUErCK4WcpQ6eMY3IiIyKQsOViao52pr2kNYbWCpGYPhRBCQEaRkZGIiIjAwoULAQAajQaBgYEYPnw43nvvvRLte/fujXv37uGHH37QHWvTpg1CQ0OxdOlSCCHg6+uLsWPHYsyYMQCAvLw81K1bF8uXL8err776xJry8vLg7u6O3NxcuLm5meibEhERUbmp1cCuXUBWFuDjA0RHA0ql5bSrIKnZo5rJP9lICQkJGDBgAFq1aoWIiAjMmzcPBQUFiI+PBwD0798ffn5+SEpKAgCMGjUKbdu2xezZs9GpUyesWbMGhw4dwmeffQYAUCgUGD16ND766COEhIQgODgYEydOhK+vL7p16ybX1yQiIqKKUCqBdu0st10lkT249e7dGzdu3MCkSZOQnZ2Nli1bYvPmzfDy8gIApKenw87OTte+TZs2WLVqFT788EO8//77CAkJwYYNG9CsWTNdm/Hjx6OgoABvvvkmcnJy8Oyzz2Lz5s1wcnKq9O9HREREZCqyD5VaIg6VEhERUWWSmj3sSj1DRERERBaFwY2IiIjISjC4EREREVkJBjciIiIiKyH7qlJLVLxeIy8vT+ZKiIiIyBYUZ44nrRllcDPg7t27AICAgACZKyEiIiJbcvfuXbi7u5d6ntuBGKDRaHD16lW4urpCoVAY9d68vDwEBAQgIyODW4lYAN4Py8N7Ynl4TywL74flqYx7IoTA3bt34evrq7d/7ePY42aAnZ0d/P39K3QNNzc3/oOzILwflof3xPLwnlgW3g/LY+57UlZPWzEuTiAiIiKyEgxuRERERFaCwc3EHB0dkZiYCEdHR7lLIfB+WCLeE8vDe2JZeD8sjyXdEy5OICIiIrIS7HEjIiIishIMbkRERERWgsGNiIiIyEowuBERERFZCQY3E1q8eDFUKhWcnJwQGRmJAwcOyF2Szfj111/RuXNn+Pr6QqFQYMOGDXrnhRCYNGkSfHx8UL16dcTExOD8+fPyFGsDkpKSEB4eDldXV9StWxfdunXD2bNn9drcv38fw4YNQ+3ateHi4oJXXnkF165dk6niqm/JkiUIDQ3VbSAaFRWFn3/+WXee90Ne06dPh0KhwOjRo3XHeE8q1+TJk6FQKPRejRs31p23lPvB4GYia9euRUJCAhITE3HkyBG0aNECsbGxuH79utyl2YSCggK0aNECixcvNnh+5syZWLBgAZYuXYr9+/fD2dkZsbGxuH//fiVXaht27tyJYcOGYd++fdi2bRsePHiADh06oKCgQNfmnXfewQ8//IBvv/0WO3fuxNWrVxEXFydj1VWbv78/pk+fjsOHD+PQoUP45z//ia5du+LUqVMAeD/kdPDgQXz66acIDQ3VO857UvmaNm2KrKws3eu3337TnbOY+yHIJCIiIsSwYcN0P6vVauHr6yuSkpJkrMo2ARDr16/X/azRaIS3t7f45JNPdMdycnKEo6OjWL16tQwV2p7r168LAGLnzp1CCO3v397eXnz77be6NqdPnxYAxN69e+Uq0+Z4eHiIL774gvdDRnfv3hUhISFi27Ztom3btmLUqFFCCP4bkUNiYqJo0aKFwXOWdD/Y42YCRUVFOHz4MGJiYnTH7OzsEBMTg71798pYGQFAWloasrOz9e6Pu7s7IiMjeX8qSW5uLgCgVq1aAIDDhw/jwYMHevekcePGCAwM5D2pBGq1GmvWrEFBQQGioqJ4P2Q0bNgwdOrUSe93D/DfiFzOnz8PX19f1KtXD/369UN6ejoAy7offMi8Cdy8eRNqtRpeXl56x728vHDmzBmZqqJi2dnZAGDw/hSfI/PRaDQYPXo0nnnmGTRr1gyA9p44ODigZs2aem15T8zrxIkTiIqKwv379+Hi4oL169ejSZMmOHbsGO+HDNasWYMjR47g4MGDJc7x30jli4yMxPLly9GoUSNkZWVhypQpiI6OxsmTJy3qfjC4EZFZDRs2DCdPntSbK0LyaNSoEY4dO4bc3FysW7cOAwYMwM6dO+UuyyZlZGRg1KhR2LZtG5ycnOQuhwC8+OKLur+HhoYiMjISQUFB+Oabb1C9enUZK9PHoVIT8PT0hFKpLLG65Nq1a/D29papKipWfA94fyrf8OHD8eOPPyI1NRX+/v66497e3igqKkJOTo5ee94T83JwcECDBg0QFhaGpKQktGjRAvPnz+f9kMHhw4dx/fp1PP3006hWrRqqVauGnTt3YsGCBahWrRq8vLx4T2RWs2ZNNGzYEBcuXLCofyMMbibg4OCAsLAwpKSk6I5pNBqkpKQgKipKxsoIAIKDg+Ht7a13f/Ly8rB//37eHzMRQmD48OFYv349fvnlFwQHB+udDwsLg729vd49OXv2LNLT03lPKpFGo0FhYSHvhwzat2+PEydO4NixY7pXq1at0K9fP93feU/klZ+fj4sXL8LHx8ei/o1wqNREEhISMGDAALRq1QoRERGYN28eCgoKEB8fL3dpNiE/Px8XLlzQ/ZyWloZjx46hVq1aCAwMxOjRo/HRRx8hJCQEwcHBmDhxInx9fdGtWzf5iq7Chg0bhlWrVuH777+Hq6urbg6Iu7s7qlevDnd3dwwaNAgJCQmoVasW3NzcMGLECERFRaF169YyV181TZgwAS+++CICAwNx9+5drFq1Cjt27MCWLVt4P2Tg6uqqm/NZzNnZGbVr19Yd5z2pXGPHjkXnzp0RFBSEq1evIjExEUqlEn369LGsfyOVuoa1ilu4cKEIDAwUDg4OIiIiQuzbt0/ukmxGamqqAFDiNWDAACGEdkuQiRMnCi8vL+Ho6Cjat28vzp49K2/RVZihewFAfPXVV7o2f/31l3j77beFh4eHqFGjhujevbvIysqSr+gq7o033hBBQUHCwcFB1KlTR7Rv315s3bpVd573Q36PbgciBO9JZevdu7fw8fERDg4Ows/PT/Tu3VtcuHBBd95S7odCCCEqNyoSERERUXlwjhsRERGRlWBwIyIiIrISDG5EREREVoLBjYiIiMhKMLgRERERWQkGNyIiIiIrweBGREREZCUY3IiIiIisBIMbEVUalUqFefPmmex6AwcONPljy3bs2AGFQlHiYdJERJaAwY2IjDZw4EAoFAooFAo4ODigQYMGmDp1Kh4+fFjm+w4ePIg333zTZHXMnz8fy5cvN9n1jHH06FH07NkTXl5ecHJyQkhICAYPHoxz587JUo+lMnVYJ7J1DG5EVC4dO3ZEVlYWzp8/jzFjxmDy5Mn45JNPDLYtKioCANSpUwc1atQwWQ3u7u6oWbOmya4n1Y8//ojWrVujsLAQK1euxOnTp/Hf//4X7u7umDhxYqXXQ0S2g8GNiMrF0dER3t7eCAoKwtChQxETE4ONGzcC+HsI8+OPP4avry8aNWoEoGTvi0KhwBdffIHu3bujRo0aCAkJ0V2j2KlTp/Dyyy/Dzc0Nrq6uiI6OxsWLF/U+p1i7du0wfPhwDB8+HO7u7vD09MTEiRPx6COZv/76a7Rq1Qqurq7w9vZG3759cf36dcnf+969e4iPj8dLL72EjRs3IiYmBsHBwYiMjMSsWbPw6aef6tru3LkTERERcHR0hI+PD9577z29Xsl27dphxIgRGD16NDw8PODl5YXPP/8cBQUFiI+Ph6urKxo0aICff/5Z957iodxNmzYhNDQUTk5OaN26NU6ePKlX53fffYemTZvC0dERKpUKs2fP1juvUqnw73//G2+88QZcXV0RGBiIzz77TK9NRkYGevXqhZo1a6JWrVro2rUrLl26pDtf/PufNWsWfHx8ULt2bQwbNgwPHjzQfb/Lly/jnXfe0fXQElHFMLgRkUlUr15d17MGACkpKTh79iy2bduGH3/8sdT3TZkyBb169cLvv/+Ol156Cf369cPt27cBAFeuXMFzzz0HR0dH/PLLLzh8+DDeeOONModkV6xYgWrVquHAgQOYP38+5syZgy+++EJ3/sGDB5g2bRqOHz+ODRs24NKlSxg4cKDk77llyxbcvHkT48ePN3i+uAfwypUreOmllxAeHo7jx49jyZIl+PLLL/HRRx+VqNfT0xMHDhzAiBEjMHToUPTs2RNt2rTBkSNH0KFDB7z++uu4d++e3vvGjRuH2bNn4+DBg6hTpw46d+6sC0yHDx9Gr1698Oqrr+LEiROYPHkyJk6cWGJYefbs2WjVqhWOHj2Kt99+G0OHDsXZs2d1v6fY2Fi4urpi165d2L17N1xcXNCxY0e9+5yamoqLFy8iNTUVK1aswPLly3Wfk5ycDH9/f0ydOhVZWVnIysqS/HsmolIIIiIjDRgwQHTt2lUIIYRGoxHbtm0Tjo6OYuzYsbrzXl5eorCwUO99QUFBYu7cubqfAYgPP/xQ93N+fr4AIH7++WchhBATJkwQwcHBoqio6Il1CCFE27ZtxVNPPSU0Go3u2LvvviueeuqpUr/LwYMHBQBx9+5dIYQQqampAoC4c+eOwfYzZswQAMTt27dLvaYQQrz//vuiUaNGerUsXrxYuLi4CLVarav32Wef1Z1/+PChcHZ2Fq+//rruWFZWlgAg9u7dq1ffmjVrdG1u3bolqlevLtauXSuEEKJv377ihRde0Ktn3LhxokmTJrqfg4KCxGuvvab7WaPRiLp164olS5YIIYT4+uuvS9RfWFgoqlevLrZs2SKE0P7+g4KCxMOHD3VtevbsKXr37q33OY/ecyKqGPa4EVG5/Pjjj3BxcYGTkxNefPFF9O7dG5MnT9adb968ORwcHJ54ndDQUN3fnZ2d4ebmphu6PHbsGKKjo2Fvby+5rtatW+sNyUVFReH8+fNQq9UAtL1RnTt3RmBgIFxdXdG2bVsAQHp6uqTri0eGXcty+vRpREVF6dXyzDPPID8/H5mZmbpjj35/pVKJ2rVro3nz5rpjXl5eAFBiODcqKkr391q1aqFRo0Y4ffq07rOfeeYZvfbPPPOM3u/h8c9WKBTw9vbWfc7x48dx4cIFuLq6wsXFBS4uLqhVqxbu37+vG6oGgKZNm0KpVOp+9vHxMWromYiMU03uAojIOj3//PNYsmQJHBwc4Ovri2rV9P9z4uzsLOk6j4cyhUIBjUYDQDv8akoFBQWIjY1FbGwsVq5ciTp16iA9PR2xsbF6w39ladiwIQDgzJkzeuGpvAx9/0ePFQe/4t+JKZX1u8/Pz0dYWBhWrlxZ4n116tSRdA0iMj32uBFRuTg7O6NBgwYIDAwsEdpMJTQ0FLt27dLN3ZJi//79ej/v27cPISEhUCqVOHPmDG7duoXp06cjOjoajRs3Nrp3qEOHDvD09MTMmTMNni/e/+2pp57C3r179Xrodu/eDVdXV/j7+xv1mYbs27dP9/c7d+7g3LlzeOqpp3SfvXv3br32u3fvRsOGDfV6x8ry9NNP4/z586hbty4aNGig93J3d5dcp4ODg14vHxFVDIMbEVms4cOHIy8vD6+++ioOHTqE8+fP4+uvv9ZNoDckPT0dCQkJOHv2LFavXo2FCxdi1KhRAIDAwEA4ODhg4cKF+PPPP7Fx40ZMmzbNqJqcnZ3xxRdfYNOmTejSpQu2b9+OS5cu4dChQxg/fjzeeustAMDbb7+NjIwMjBgxAmfOnMH333+PxMREJCQkwM6u4v/pnTp1KlJSUnDy5EkMHDgQnp6euhW2Y8aMQUpKCqZNm4Zz585hxYoVWLRoEcaOHSv5+v369YOnpye6du2KXbt2IS0tDTt27MDIkSP1hnqfRKVS4ddff8WVK1dw8+ZNY78mET2GwY2ILFbt2rXxyy+/ID8/H23btkVYWBg+//zzMue89e/fH3/99RciIiIwbNgwjBo1Srfpb506dbB8+XJ8++23aNKkCaZPn45Zs2YZXVfXrl2xZ88e2Nvbo2/fvmjcuDH69OmD3Nxc3apRPz8//PTTTzhw4ABatGiBt956C4MGDcKHH35Yvl/GY6ZPn45Ro0YhLCwM2dnZ+OGHH3RzCp9++ml88803WLNmDZo1a4ZJkyZh6tSpRq2erVGjBn799VcEBgYiLi4OTz31FAYNGoT79+/Dzc1N8nWmTp2KS5cuoX79+npDrERUPgohdaYtEZGFa9euHVq2bFmld+rfsWMHnn/+edy5c0eWzYeJSF7scSMiIiKyEgxuRERERFaCQ6VEREREVoI9bkRERERWgsGNiIiIyEowuBERERFZCQY3IiIiIivB4EZERERkJRjciIiIiKwEgxsRERGRlWBwIyIiIrIS/w8CmZv6eY814QAAAABJRU5ErkJggg=="
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![num_comp.png](attachment:num_comp.png)        .   .     ![scree_plot.png](attachment:scree_plot.png)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The plots above shows the cumulative variance increasing the number of principal components generated by PCA procedure.\n",
    "Running our algorithms we noticed that, using principal component data on our classification algorithms the performances becomes worse. This may be due to the fact that during the process of dimensionality reduction, PCA may result in the loss of detailed information on individual columns of the original data. This happens because PCA operates on the basis of data variance and tries to reduce complexity by retaining only the most relevant information. The principal components identified by PCA are linear combinations of the original columns, so some column-specific information may be merged together or even eliminated in the projection process./\n",
    "Due to this we decided to do not implement a PCA and use a model agnostic ***Permutation Importance*** procedure (as described in Elements of SL,chapter \"15.3.2 Variable Importance\") to filter our data and select the most important features"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Resampling with SMOTE"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The data we are dealing with are strongly unbalanced with regard to the classes to be classified:\n",
    "\n",
    "Label 1: 0.935143% of data;\\\n",
    "Label 2: 0.058824;\\\n",
    "Label 3: 0.006033.\n",
    "\n",
    "To overcame this issue we have tried to use a SMOTE tecnique of over/under sampling.\\\n",
    "Please furthermore notice that, after a data exploration we noticed that:\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>main.pos</th>\n",
       "      <th>fun.pos</th>\n",
       "      <th>when</th>\n",
       "      <th>y</th>\n",
       "      <th>E1A1</th>\n",
       "      <th>E1A2</th>\n",
       "      <th>E1A3</th>\n",
       "      <th>E1A4</th>\n",
       "      <th>E1A5</th>\n",
       "      <th>...</th>\n",
       "      <th>M2V2041</th>\n",
       "      <th>M2V2042</th>\n",
       "      <th>M2V2043</th>\n",
       "      <th>M2V2044</th>\n",
       "      <th>M2V2045</th>\n",
       "      <th>M2V2046</th>\n",
       "      <th>M2V2047</th>\n",
       "      <th>M2V2048</th>\n",
       "      <th>Season</th>\n",
       "      <th>elapsed_time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>29</td>\n",
       "      <td>LHCB-6</td>\n",
       "      <td>QSCB-6-CP7</td>\n",
       "      <td>2018-06-01</td>\n",
       "      <td>3</td>\n",
       "      <td>3.849429</td>\n",
       "      <td>-1.120215</td>\n",
       "      <td>1.900126</td>\n",
       "      <td>1.765498</td>\n",
       "      <td>-3.954946</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.035500</td>\n",
       "      <td>-0.032582</td>\n",
       "      <td>0.158069</td>\n",
       "      <td>0.152430</td>\n",
       "      <td>0.453890</td>\n",
       "      <td>0.563897</td>\n",
       "      <td>0.683447</td>\n",
       "      <td>-0.023237</td>\n",
       "      <td>2</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>332</th>\n",
       "      <td>514</td>\n",
       "      <td>LHCB-6</td>\n",
       "      <td>QSCB-6-CP7</td>\n",
       "      <td>2017-11-01</td>\n",
       "      <td>3</td>\n",
       "      <td>-1.329898</td>\n",
       "      <td>-1.862015</td>\n",
       "      <td>-1.651927</td>\n",
       "      <td>-1.171916</td>\n",
       "      <td>0.036863</td>\n",
       "      <td>...</td>\n",
       "      <td>0.003700</td>\n",
       "      <td>-0.093183</td>\n",
       "      <td>0.167388</td>\n",
       "      <td>0.251590</td>\n",
       "      <td>0.381014</td>\n",
       "      <td>0.320008</td>\n",
       "      <td>0.095782</td>\n",
       "      <td>-0.094574</td>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>568</th>\n",
       "      <td>874</td>\n",
       "      <td>LHCB-6</td>\n",
       "      <td>QSCB-6-CP7</td>\n",
       "      <td>2017-12-01</td>\n",
       "      <td>3</td>\n",
       "      <td>1.988860</td>\n",
       "      <td>0.591283</td>\n",
       "      <td>1.497192</td>\n",
       "      <td>0.372399</td>\n",
       "      <td>1.557347</td>\n",
       "      <td>...</td>\n",
       "      <td>0.296484</td>\n",
       "      <td>0.025949</td>\n",
       "      <td>-0.071119</td>\n",
       "      <td>-0.088406</td>\n",
       "      <td>0.046136</td>\n",
       "      <td>-0.029481</td>\n",
       "      <td>0.014090</td>\n",
       "      <td>-0.267449</td>\n",
       "      <td>4</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>575</th>\n",
       "      <td>885</td>\n",
       "      <td>LHCB-8</td>\n",
       "      <td>QSCB-8-CP1</td>\n",
       "      <td>2020-10-01</td>\n",
       "      <td>3</td>\n",
       "      <td>-0.685269</td>\n",
       "      <td>0.234567</td>\n",
       "      <td>-0.149447</td>\n",
       "      <td>-0.076372</td>\n",
       "      <td>-0.050782</td>\n",
       "      <td>...</td>\n",
       "      <td>0.053421</td>\n",
       "      <td>0.123710</td>\n",
       "      <td>0.063862</td>\n",
       "      <td>0.047044</td>\n",
       "      <td>0.082810</td>\n",
       "      <td>-0.020352</td>\n",
       "      <td>0.063948</td>\n",
       "      <td>0.054148</td>\n",
       "      <td>3</td>\n",
       "      <td>45</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4 rows × 49159 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      id main.pos     fun.pos       when  y      E1A1      E1A2      E1A3  \\\n",
       "19    29   LHCB-6  QSCB-6-CP7 2018-06-01  3  3.849429 -1.120215  1.900126   \n",
       "332  514   LHCB-6  QSCB-6-CP7 2017-11-01  3 -1.329898 -1.862015 -1.651927   \n",
       "568  874   LHCB-6  QSCB-6-CP7 2017-12-01  3  1.988860  0.591283  1.497192   \n",
       "575  885   LHCB-8  QSCB-8-CP1 2020-10-01  3 -0.685269  0.234567 -0.149447   \n",
       "\n",
       "         E1A4      E1A5  ...   M2V2041   M2V2042   M2V2043   M2V2044  \\\n",
       "19   1.765498 -3.954946  ... -0.035500 -0.032582  0.158069  0.152430   \n",
       "332 -1.171916  0.036863  ...  0.003700 -0.093183  0.167388  0.251590   \n",
       "568  0.372399  1.557347  ...  0.296484  0.025949 -0.071119 -0.088406   \n",
       "575 -0.076372 -0.050782  ...  0.053421  0.123710  0.063862  0.047044   \n",
       "\n",
       "      M2V2045   M2V2046   M2V2047   M2V2048  Season  elapsed_time  \n",
       "19   0.453890  0.563897  0.683447 -0.023237       2            17  \n",
       "332  0.381014  0.320008  0.095782 -0.094574       3            10  \n",
       "568  0.046136 -0.029481  0.014090 -0.267449       4            11  \n",
       "575  0.082810 -0.020352  0.063948  0.054148       3            45  \n",
       "\n",
       "[4 rows x 49159 columns]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.y==3]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "three maximum alert level (label 3) have been detected by the same main position of sensors in the LHC. Generally LHCB main position seems to be a high risk position for \"Big fail\" alert level."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "def resampling_strategy(train, y):\n",
    "    counts = y.value_counts()\n",
    "    total_new_train = int(np.ceil(counts[1]*100/70))\n",
    "    sampling_strategy={2: int(np.ceil(total_new_train*0.25)), 3: int(np.floor(total_new_train*0.05))}\n",
    "    return sampling_strategy\n",
    "\n",
    "#smote = SMOTE(sampling_strategy=resampling_strategy(principalDf, df.y), k_neighbors=3)\n",
    "smote = SMOTE(k_neighbors=3)#sampling_strategy=resampling_strategy(X_tr_scaled, df.y), k_neighbors=3)\n",
    "X_train_resampled, y_train_resampled = smote.fit_resample(X_tr_scaled, df.y)  "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Feature importance"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Find Variable Importance using Permutation importance as described in Elememnts Of SL book. Notice that we use a sort of cross validation on the different classifiers and compute the mean of all the importance scores. "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### PERMUTATION IMPORTANCE "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The main idea behind Permutation Importance is to measure how much the predictive performance of a model deteriorates when the values of a particular feature are randomly permuted. The Permutation Importance scores can be used to rank the features based on their importance. Features with higher scores are considered more influential in the model's predictions, while features with lower scores are deemed less important."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.inspection import permutation_importance\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "#PERMUTATION IMPORTANCE WITHout CROSS VALIDATION \n",
    "X_train, X_val, y_train, y_val = train_test_split(X_train_resampled,y_train_resampled)\n",
    "model = LogisticRegression(multi_class='multinomial', class_weight=\"balanced\",max_iter=1000,n_jobs=-1)\n",
    "lr = model.fit(X_train, y_train-1)\n",
    "\n",
    "r = permutation_importance(model, X_val, y_val, n_repeats=100, n_jobs=-1, random_state=0)\n",
    "importance_df = pd.DataFrame()\n",
    "importance_df[\"Features\"] = X_tr_scaled.columns\n",
    "importance_df['Importances'] = r.importances_mean\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>E1A_mag_max_6</th>\n",
       "      <th>E1V_mag_max_7</th>\n",
       "      <th>E1V_mag_max_8</th>\n",
       "      <th>E2A_mag_max_11</th>\n",
       "      <th>E2H_mag_max_8</th>\n",
       "      <th>E2V_mag_max_8</th>\n",
       "      <th>E3H_mag_max_10</th>\n",
       "      <th>E3H_mag_max_11</th>\n",
       "      <th>E4A_mag_max_6</th>\n",
       "      <th>E4A_mag_max_7</th>\n",
       "      <th>...</th>\n",
       "      <th>M1H_mag_max_7</th>\n",
       "      <th>M1H_mag_max_8</th>\n",
       "      <th>M1H_mag_max_1</th>\n",
       "      <th>M1H_mag_max_5</th>\n",
       "      <th>M1V_mag_max_10</th>\n",
       "      <th>M2A_mag_max_1</th>\n",
       "      <th>M2H_mag_max_5</th>\n",
       "      <th>M2V_mag_max_4</th>\n",
       "      <th>E2A_psd2</th>\n",
       "      <th>cor_E3V_E4A</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.039422</td>\n",
       "      <td>7.020925</td>\n",
       "      <td>15.585273</td>\n",
       "      <td>0.382115</td>\n",
       "      <td>-0.142112</td>\n",
       "      <td>0.661720</td>\n",
       "      <td>0.586508</td>\n",
       "      <td>0.428370</td>\n",
       "      <td>0.619162</td>\n",
       "      <td>-0.105400</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.270466</td>\n",
       "      <td>-0.226682</td>\n",
       "      <td>-0.125634</td>\n",
       "      <td>-0.261951</td>\n",
       "      <td>-0.143528</td>\n",
       "      <td>-0.294740</td>\n",
       "      <td>-0.179505</td>\n",
       "      <td>2.364448</td>\n",
       "      <td>-0.193553</td>\n",
       "      <td>1.120693</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.028135</td>\n",
       "      <td>-0.143063</td>\n",
       "      <td>-0.165264</td>\n",
       "      <td>0.108456</td>\n",
       "      <td>0.328507</td>\n",
       "      <td>0.092942</td>\n",
       "      <td>-0.268614</td>\n",
       "      <td>-0.187020</td>\n",
       "      <td>0.395465</td>\n",
       "      <td>-0.034374</td>\n",
       "      <td>...</td>\n",
       "      <td>4.160368</td>\n",
       "      <td>1.524481</td>\n",
       "      <td>1.535556</td>\n",
       "      <td>1.891048</td>\n",
       "      <td>0.309241</td>\n",
       "      <td>0.143925</td>\n",
       "      <td>2.432722</td>\n",
       "      <td>1.347164</td>\n",
       "      <td>-0.133627</td>\n",
       "      <td>-2.388895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.070351</td>\n",
       "      <td>-0.136579</td>\n",
       "      <td>-0.152090</td>\n",
       "      <td>-0.016347</td>\n",
       "      <td>3.479717</td>\n",
       "      <td>0.120743</td>\n",
       "      <td>-0.440804</td>\n",
       "      <td>-0.403386</td>\n",
       "      <td>-0.185856</td>\n",
       "      <td>0.019921</td>\n",
       "      <td>...</td>\n",
       "      <td>6.321317</td>\n",
       "      <td>3.475789</td>\n",
       "      <td>4.058494</td>\n",
       "      <td>0.089018</td>\n",
       "      <td>-0.067613</td>\n",
       "      <td>0.000568</td>\n",
       "      <td>0.682820</td>\n",
       "      <td>2.251462</td>\n",
       "      <td>0.700701</td>\n",
       "      <td>0.408244</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.705323</td>\n",
       "      <td>0.007500</td>\n",
       "      <td>-0.125810</td>\n",
       "      <td>0.395393</td>\n",
       "      <td>0.971524</td>\n",
       "      <td>1.242974</td>\n",
       "      <td>0.269326</td>\n",
       "      <td>0.837783</td>\n",
       "      <td>-0.039785</td>\n",
       "      <td>0.310685</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.217263</td>\n",
       "      <td>-0.294760</td>\n",
       "      <td>-0.148706</td>\n",
       "      <td>-0.207138</td>\n",
       "      <td>-0.142485</td>\n",
       "      <td>0.527448</td>\n",
       "      <td>-0.267823</td>\n",
       "      <td>-0.341132</td>\n",
       "      <td>-0.085938</td>\n",
       "      <td>0.178487</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.041541</td>\n",
       "      <td>6.471077</td>\n",
       "      <td>11.459325</td>\n",
       "      <td>-0.243782</td>\n",
       "      <td>-0.181382</td>\n",
       "      <td>-0.213556</td>\n",
       "      <td>-0.420938</td>\n",
       "      <td>-0.276111</td>\n",
       "      <td>-0.144902</td>\n",
       "      <td>-0.076566</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.277995</td>\n",
       "      <td>-0.314340</td>\n",
       "      <td>0.785985</td>\n",
       "      <td>-0.351034</td>\n",
       "      <td>-0.046554</td>\n",
       "      <td>0.315969</td>\n",
       "      <td>-0.246636</td>\n",
       "      <td>-0.320740</td>\n",
       "      <td>-0.262387</td>\n",
       "      <td>-0.504943</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1855</th>\n",
       "      <td>-0.050005</td>\n",
       "      <td>2.473184</td>\n",
       "      <td>1.160485</td>\n",
       "      <td>2.357049</td>\n",
       "      <td>7.728049</td>\n",
       "      <td>2.770251</td>\n",
       "      <td>0.517038</td>\n",
       "      <td>0.840202</td>\n",
       "      <td>0.118665</td>\n",
       "      <td>0.388926</td>\n",
       "      <td>...</td>\n",
       "      <td>4.755698</td>\n",
       "      <td>1.878494</td>\n",
       "      <td>2.062005</td>\n",
       "      <td>0.915332</td>\n",
       "      <td>0.175444</td>\n",
       "      <td>0.635714</td>\n",
       "      <td>3.011255</td>\n",
       "      <td>1.445486</td>\n",
       "      <td>1.425179</td>\n",
       "      <td>-0.396976</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1856</th>\n",
       "      <td>-0.086198</td>\n",
       "      <td>-0.009324</td>\n",
       "      <td>0.205164</td>\n",
       "      <td>-0.036945</td>\n",
       "      <td>1.808775</td>\n",
       "      <td>0.649714</td>\n",
       "      <td>-0.402551</td>\n",
       "      <td>-0.128795</td>\n",
       "      <td>0.200215</td>\n",
       "      <td>0.054089</td>\n",
       "      <td>...</td>\n",
       "      <td>1.590554</td>\n",
       "      <td>0.797462</td>\n",
       "      <td>0.077092</td>\n",
       "      <td>1.304962</td>\n",
       "      <td>-0.011572</td>\n",
       "      <td>2.336294</td>\n",
       "      <td>0.285787</td>\n",
       "      <td>1.717604</td>\n",
       "      <td>0.054950</td>\n",
       "      <td>-0.158217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1857</th>\n",
       "      <td>-0.014505</td>\n",
       "      <td>-0.023813</td>\n",
       "      <td>0.751722</td>\n",
       "      <td>0.261055</td>\n",
       "      <td>4.807867</td>\n",
       "      <td>1.854874</td>\n",
       "      <td>-0.349493</td>\n",
       "      <td>0.098330</td>\n",
       "      <td>0.576282</td>\n",
       "      <td>-0.073375</td>\n",
       "      <td>...</td>\n",
       "      <td>4.287171</td>\n",
       "      <td>2.508887</td>\n",
       "      <td>0.247987</td>\n",
       "      <td>3.884981</td>\n",
       "      <td>0.184465</td>\n",
       "      <td>0.849475</td>\n",
       "      <td>1.114234</td>\n",
       "      <td>4.913256</td>\n",
       "      <td>0.752987</td>\n",
       "      <td>-0.053407</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1858</th>\n",
       "      <td>14.627397</td>\n",
       "      <td>1.478908</td>\n",
       "      <td>2.665996</td>\n",
       "      <td>3.101822</td>\n",
       "      <td>1.035617</td>\n",
       "      <td>1.464679</td>\n",
       "      <td>6.111525</td>\n",
       "      <td>7.257406</td>\n",
       "      <td>5.069790</td>\n",
       "      <td>22.357813</td>\n",
       "      <td>...</td>\n",
       "      <td>0.683273</td>\n",
       "      <td>0.065317</td>\n",
       "      <td>0.484813</td>\n",
       "      <td>1.274782</td>\n",
       "      <td>1.905680</td>\n",
       "      <td>2.257431</td>\n",
       "      <td>0.510563</td>\n",
       "      <td>0.445573</td>\n",
       "      <td>2.441543</td>\n",
       "      <td>-0.678621</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1859</th>\n",
       "      <td>30.577306</td>\n",
       "      <td>3.972631</td>\n",
       "      <td>6.109061</td>\n",
       "      <td>7.597022</td>\n",
       "      <td>4.766842</td>\n",
       "      <td>4.085355</td>\n",
       "      <td>13.517489</td>\n",
       "      <td>15.762996</td>\n",
       "      <td>10.599107</td>\n",
       "      <td>46.446396</td>\n",
       "      <td>...</td>\n",
       "      <td>3.027831</td>\n",
       "      <td>1.048501</td>\n",
       "      <td>1.781857</td>\n",
       "      <td>3.184009</td>\n",
       "      <td>4.200795</td>\n",
       "      <td>0.355898</td>\n",
       "      <td>2.409316</td>\n",
       "      <td>1.554188</td>\n",
       "      <td>6.077090</td>\n",
       "      <td>-1.242118</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1860 rows × 27 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      E1A_mag_max_6  E1V_mag_max_7  E1V_mag_max_8  E2A_mag_max_11  \\\n",
       "0          0.039422       7.020925      15.585273        0.382115   \n",
       "1          0.028135      -0.143063      -0.165264        0.108456   \n",
       "2         -0.070351      -0.136579      -0.152090       -0.016347   \n",
       "3          0.705323       0.007500      -0.125810        0.395393   \n",
       "4         -0.041541       6.471077      11.459325       -0.243782   \n",
       "...             ...            ...            ...             ...   \n",
       "1855      -0.050005       2.473184       1.160485        2.357049   \n",
       "1856      -0.086198      -0.009324       0.205164       -0.036945   \n",
       "1857      -0.014505      -0.023813       0.751722        0.261055   \n",
       "1858      14.627397       1.478908       2.665996        3.101822   \n",
       "1859      30.577306       3.972631       6.109061        7.597022   \n",
       "\n",
       "      E2H_mag_max_8  E2V_mag_max_8  E3H_mag_max_10  E3H_mag_max_11  \\\n",
       "0         -0.142112       0.661720        0.586508        0.428370   \n",
       "1          0.328507       0.092942       -0.268614       -0.187020   \n",
       "2          3.479717       0.120743       -0.440804       -0.403386   \n",
       "3          0.971524       1.242974        0.269326        0.837783   \n",
       "4         -0.181382      -0.213556       -0.420938       -0.276111   \n",
       "...             ...            ...             ...             ...   \n",
       "1855       7.728049       2.770251        0.517038        0.840202   \n",
       "1856       1.808775       0.649714       -0.402551       -0.128795   \n",
       "1857       4.807867       1.854874       -0.349493        0.098330   \n",
       "1858       1.035617       1.464679        6.111525        7.257406   \n",
       "1859       4.766842       4.085355       13.517489       15.762996   \n",
       "\n",
       "      E4A_mag_max_6  E4A_mag_max_7  ...  M1H_mag_max_7  M1H_mag_max_8  \\\n",
       "0          0.619162      -0.105400  ...      -0.270466      -0.226682   \n",
       "1          0.395465      -0.034374  ...       4.160368       1.524481   \n",
       "2         -0.185856       0.019921  ...       6.321317       3.475789   \n",
       "3         -0.039785       0.310685  ...      -0.217263      -0.294760   \n",
       "4         -0.144902      -0.076566  ...      -0.277995      -0.314340   \n",
       "...             ...            ...  ...            ...            ...   \n",
       "1855       0.118665       0.388926  ...       4.755698       1.878494   \n",
       "1856       0.200215       0.054089  ...       1.590554       0.797462   \n",
       "1857       0.576282      -0.073375  ...       4.287171       2.508887   \n",
       "1858       5.069790      22.357813  ...       0.683273       0.065317   \n",
       "1859      10.599107      46.446396  ...       3.027831       1.048501   \n",
       "\n",
       "      M1H_mag_max_1  M1H_mag_max_5  M1V_mag_max_10  M2A_mag_max_1  \\\n",
       "0         -0.125634      -0.261951       -0.143528      -0.294740   \n",
       "1          1.535556       1.891048        0.309241       0.143925   \n",
       "2          4.058494       0.089018       -0.067613       0.000568   \n",
       "3         -0.148706      -0.207138       -0.142485       0.527448   \n",
       "4          0.785985      -0.351034       -0.046554       0.315969   \n",
       "...             ...            ...             ...            ...   \n",
       "1855       2.062005       0.915332        0.175444       0.635714   \n",
       "1856       0.077092       1.304962       -0.011572       2.336294   \n",
       "1857       0.247987       3.884981        0.184465       0.849475   \n",
       "1858       0.484813       1.274782        1.905680       2.257431   \n",
       "1859       1.781857       3.184009        4.200795       0.355898   \n",
       "\n",
       "      M2H_mag_max_5  M2V_mag_max_4  E2A_psd2  cor_E3V_E4A  \n",
       "0         -0.179505       2.364448 -0.193553     1.120693  \n",
       "1          2.432722       1.347164 -0.133627    -2.388895  \n",
       "2          0.682820       2.251462  0.700701     0.408244  \n",
       "3         -0.267823      -0.341132 -0.085938     0.178487  \n",
       "4         -0.246636      -0.320740 -0.262387    -0.504943  \n",
       "...             ...            ...       ...          ...  \n",
       "1855       3.011255       1.445486  1.425179    -0.396976  \n",
       "1856       0.285787       1.717604  0.054950    -0.158217  \n",
       "1857       1.114234       4.913256  0.752987    -0.053407  \n",
       "1858       0.510563       0.445573  2.441543    -0.678621  \n",
       "1859       2.409316       1.554188  6.077090    -1.242118  \n",
       "\n",
       "[1860 rows x 27 columns]"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_tr = X_train_resampled[importance_df[importance_df[\"Importances\"] > 0.0002].Features.values]\n",
    "X_tr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#no correlation inside the  best features!\n",
    "X_tr.columns.str.contains(\"cor\").sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "#X_tr.to_pickle(\"X_56_COLS\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# CV for logistic regression"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Cross Validation to choose the best parameters:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n",
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:778: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 767, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 234, in __call__\n",
      "    return self._score(\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\metrics\\_scorer.py\", line 282, in _score\n",
      "    return self._sign * self._score_func(y_true, y_pred, **self._kwargs)\n",
      "  File \"C:\\Users\\Admin\\AppData\\Local\\Temp\\ipykernel_9248\\2198874025.py\", line 26, in bas\n",
      "    y_true_label = y_true[:, i]\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 984, in __getitem__\n",
      "    return self._get_with(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 999, in _get_with\n",
      "    return self._get_values_tuple(key)\n",
      "  File \"c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\pandas\\core\\series.py\", line 1034, in _get_values_tuple\n",
      "    raise KeyError(\"key of type tuple not found and not a MultiIndex\")\n",
      "KeyError: 'key of type tuple not found and not a MultiIndex'\n",
      "\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "#CROSS VALIDATION\n",
    "from sklearn.model_selection import RandomizedSearchCV,StratifiedKFold\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import make_scorer,confusion_matrix\n",
    "\n",
    "tol = [1e-4, 3e-4, 1e-3, 5e-4]\n",
    "fit_intercept = [True, False]\n",
    "solver= ['lbfgs', 'liblinear', 'newton-cg', 'newton-cholesky', 'sag', 'saga']\n",
    "warm_start=[True,False]\n",
    "n_iter=[100,1000,10000]\n",
    "penalty = ['l1', 'l2']\n",
    "C = [ 0.01, 0.1, 1, 10, 100,1000]\n",
    "hyperparameters = dict(penalty=penalty, C=C,warm_start=warm_start,max_iter=n_iter,solver=solver,tol=tol,fit_intercept=fit_intercept)\n",
    "#fit the model \n",
    "model = LogisticRegression(multi_class='multinomial', max_iter=10000, class_weight=\"balanced\",n_jobs=-1)\n",
    "\n",
    "\n",
    "def bas(y_true, y_pred):\n",
    "    num_labels = 3\n",
    "    #sensitivities = []\n",
    "    #specificities = []\n",
    "    recall=[]\n",
    "    class_counts={1 :(1/0.935143),2:(1/0.058824),3 :(1/0.006033)}\n",
    "\n",
    "    for i in range(num_labels):\n",
    "        y_true_label = y_true[:, i]\n",
    "        y_pred_label = y_pred[:, i]\n",
    "        \n",
    "        tn, fp, fn, tp = confusion_matrix(y_true_label, y_pred_label).ravel()\n",
    "        #sensitivity = tp / (tp + fn)\n",
    "        #specificity = tn / (tn + fp)\n",
    "        rec= tp / (tp+fn)*class_counts.values[i]\n",
    "        \n",
    "        recall.append(rec)\n",
    "        #sensitivities.append(sensitivity)\n",
    "        #specificities.append(specificity)\n",
    "    \n",
    "    bas= np.mean(recall)\n",
    "    #mean_sensitivity = np.mean(sensitivities)\n",
    "    #mean_specificity = np.mean(specificities)\n",
    "    \n",
    "    return bas\n",
    "    #return 0.5 * (mean_sensitivity + mean_specificity)\n",
    "\n",
    "bas_scorer = make_scorer(bas)\n",
    "\n",
    "random_search = RandomizedSearchCV(model, hyperparameters, n_iter=100, cv=StratifiedKFold(n_splits=5), random_state=42, scoring=bas_scorer)\n",
    "#random_search.fit(X_train_resampled, y_train_resampled)\n",
    "random_search.fit(X_tr, y_train_resampled)\n",
    "best_params=random_search.best_params_\n",
    "print(\"Best parameters:\", random_search.best_params_) \n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols=X_tr.columns.to_list()\n",
    "X_te = X_te_scaled[cols]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2, 1, 1, 1, 1, 2, 2, 1, 2, 1, 2, 2, 2, 2, 1, 1, 1, 2, 1, 2, 1, 2,\n",
       "       1, 2, 1, 1, 2, 1, 2, 2, 2, 2, 2, 1, 2, 1, 1, 1, 1, 2, 1, 1, 2, 1,\n",
       "       1, 1, 2, 1, 1, 1, 1, 2, 2, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1,\n",
       "       1, 3, 1, 1, 1, 1, 1, 1, 1, 3, 1, 1, 1, 2, 1, 1, 1, 1, 2, 1, 1, 1,\n",
       "       1, 1, 2, 1, 1, 1, 2, 2, 2, 2, 2, 1, 2, 3, 1, 1, 2, 1, 2, 3, 2, 1,\n",
       "       1, 1, 1, 2, 1, 2, 1, 2, 1, 1, 1, 2, 2, 2, 1, 3, 1, 1, 2, 1, 2, 2,\n",
       "       1, 1, 2, 2, 1, 2, 2, 2, 3, 1, 1, 2, 1, 1, 1, 3, 1, 1, 1, 3, 1, 1,\n",
       "       1, 2, 1, 1, 1, 1, 1, 2, 2, 1, 2, 2, 1, 1, 2, 3, 1, 1, 1, 1, 2, 2,\n",
       "       1, 1, 2, 1, 2, 2, 2, 1, 2, 2, 1, 3, 2, 1, 1, 2, 1, 1, 1, 1, 1, 2,\n",
       "       1, 2, 1, 1, 1, 3, 1, 2, 1, 1, 1, 1, 2, 2, 2, 1, 2, 2, 1, 2, 1, 1,\n",
       "       2, 2, 1, 1, 3, 1, 2, 1, 1, 1, 1, 1, 2, 2, 2, 1, 1, 1, 1, 1, 1, 2,\n",
       "       2, 3, 1, 1, 1, 1, 3, 1, 1, 2, 2, 2, 3, 1, 2, 1, 2, 2, 1, 1, 2, 1,\n",
       "       1, 1, 1, 1, 2, 3, 2, 1, 1, 3, 2, 1, 1, 1, 1, 2, 1, 1, 2, 1, 2, 1,\n",
       "       1, 3, 1, 1, 1, 1, 1, 1, 3, 2, 1, 1, 1, 1, 1, 1, 2, 2, 1, 1, 2, 1,\n",
       "       2, 1, 1, 1, 1, 1, 1, 2, 2, 1, 2, 1, 2, 2, 3, 2, 1, 1, 1, 1, 1, 1,\n",
       "       1, 1, 1, 1, 2, 3, 2, 1, 3, 1, 2, 1, 1, 2, 2, 2, 1, 2, 1, 2, 1, 1,\n",
       "       2, 1, 1], dtype=int64)"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_model = random_search.best_estimator_\n",
    "y_pred = best_model.predict(X_te)\n",
    "y_pred"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Logistic Regression with weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 332,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           1       0.83      0.69      0.75       116\n",
      "           2       0.58      0.68      0.62        56\n",
      "           3       0.38      1.00      0.55         6\n",
      "\n",
      "    accuracy                           0.70       178\n",
      "   macro avg       0.59      0.79      0.64       178\n",
      "weighted avg       0.74      0.70      0.71       178\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\Admin\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import classification_report\n",
    "\n",
    "\n",
    "# Split the data into training and testing sets\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_train_resampled, y_train_resampled, test_size=0.2)\n",
    "\n",
    "# Define the class weights inversely proportional to their frequencies\n",
    "#weights = {\"1\": 1 / 620, \"2\": 1 / 39, \"3\": 1 / 4}\n",
    "\n",
    "# Create an instance of LogisticRegression for multinomial classification\n",
    "model = LogisticRegression(multi_class='multinomial', class_weight=\"balanced\",**best_params)\n",
    "\n",
    "# Train the logistic regression model\n",
    "model.fit(X_train, (y_train)-1)\n",
    "\n",
    "# Predict the target variable for the test set\n",
    "preds = model.predict(X_test)\n",
    "predictions= (preds)+1\n",
    "\n",
    "# Evaluate the model\n",
    "classification_report = classification_report(y_test, predictions)\n",
    "print(classification_report)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 333,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 1, 1, 3, 2, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
       "       2, 1, 1, 1, 2, 2, 1, 1, 1, 1, 2, 1, 2, 1, 2, 1, 1, 2, 2, 1, 3, 1,\n",
       "       2, 2, 1, 1, 2, 1, 1, 1, 1, 3, 2, 1, 2, 1, 2, 2, 1, 1, 2, 1, 1, 1,\n",
       "       1, 2, 1, 1, 1, 1, 1, 1, 2, 2, 3, 1, 1, 1, 1, 1, 1, 2, 1, 2, 1, 1,\n",
       "       1, 1, 1, 2, 2, 1, 2, 1, 2, 1, 1, 1, 1, 2, 2, 1, 1, 1, 1, 2, 1, 1,\n",
       "       1, 1, 1, 2, 2, 1, 1, 2, 1, 1, 1, 3, 1, 1, 2, 2, 2, 2, 1, 1, 1, 1,\n",
       "       1, 1, 1, 1, 1, 1, 1, 2, 2, 1, 3, 1, 1, 1, 1, 1, 2, 2, 1, 3, 1, 1,\n",
       "       1, 2, 2, 1, 2, 1, 1, 3, 2, 1, 1, 1, 1, 1, 1, 2, 2, 2, 1, 1, 2, 2,\n",
       "       2, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 2, 2, 1, 2, 1, 2, 2,\n",
       "       1, 1, 1, 1, 3, 2, 1, 2, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 2, 1, 1, 1,\n",
       "       1, 3, 1, 2, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
       "       1, 2, 2, 1, 1, 1, 1, 3, 1, 1, 2, 1, 2, 2, 1, 2, 2, 2, 1, 2, 1, 3,\n",
       "       2, 2, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 2, 2, 1, 1, 2, 1, 1, 1, 1, 1,\n",
       "       1, 2, 1, 2, 1, 1, 2, 1, 2, 1, 1, 1, 1, 1, 3, 1, 2, 1, 2, 2, 2, 2,\n",
       "       1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 3, 1, 2, 1, 1, 2, 1, 2, 1, 1,\n",
       "       3, 2, 1, 2, 1, 3, 1, 1, 2, 2, 1, 1, 1, 2, 1, 2, 2, 2, 2, 1, 2, 2,\n",
       "       1, 1, 1], dtype=int64)"
      ]
     },
     "execution_count": 333,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions=model.predict(X_te)\n",
    "predictions= (predictions)+1\n",
    "preditctions"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Consegna"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "def consegna(pretictions, i):\n",
    "    \"\"\"\n",
    "    predictions: array-like set of predictions from our model\n",
    "    i: int representing the number of the output we are uploading\n",
    "    \"\"\"\n",
    "    out = pd.DataFrame({\"id\": test_df[\"id\"], \"y\": pretictions})\n",
    "    out.to_csv(f'output{i}.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "consegna(y_pred,\"_IMPORTANT_NONIMMACOLATO\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.1"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}