Explore_downsampling.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load all neccessary packages \n",
    "import pandas as pd\n",
    "import warnings\n",
    "import numpy as np\n",
    "import sys\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "from noise_helpers.indicators import *\n",
    "from noise_helpers.labels import *\n",
    "from noise_helpers.is_class import *\n",
    "\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import f1_score\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import GroupKFold\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.metrics import r2_score\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.model_selection import GroupKFold\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib.colors import LinearSegmentedColormap\n",
    "import matplotlib.ticker as tk\n",
    "from matplotlib.ticker import PercentFormatter , NullFormatter\n",
    "from collections import defaultdict\n",
    "from sklearn.metrics import confusion_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv(\"./Data/data.csv\")\n",
    "data = data.fillna(0)\n",
    "data = data[(data[\"Leq\"] != 0)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The RMSE between the Actual and the Downsampled "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def n_samples(F):\n",
    "    N =[]\n",
    "    for f in F:\n",
    "        N.append((900 / int (f)))\n",
    "    return N\n",
    "\n",
    "rmse_values = []\n",
    "\n",
    "L_indicators = [data[\"Leq\"], data[\"L10\"], data[\"L90\"]]\n",
    "\n",
    "L4_indicators = [data[\"Leq_4\"], data[\"L10_4\"], data[\"L90_4\"]]\n",
    "L10_indicators = [data[\"Leq_10\"], data[\"L10_10\"], data[\"L90_10\"]]\n",
    "L20_indicators = [data[\"Leq_20\"], data[\"L10_20\"], data[\"L90_20\"]]\n",
    "L30_indicators = [data[\"Leq_30\"], data[\"L10_30\"], data[\"L90_30\"]]\n",
    "L60_indicators = [data[\"Leq_60\"], data[\"L10_60\"], data[\"L90_60\"]]\n",
    "L120_indicators = [data[\"Leq_120\"], data[\"L10_120\"], data[\"L90_120\"]]\n",
    "L180_indicators = [data[\"Leq_180\"], data[\"L10_180\"], data[\"L90_180\"]]\n",
    "L300_indicators = [data[\"Leq_300\"], data[\"L10_300\"], data[\"L90_300\"]]\n",
    "L450_indicators = [data[\"Leq_450\"], data[\"L10_450\"], data[\"L90_450\"]]\n",
    "L900_indicators = [data[\"Leq_900\"], data[\"L10_900\"], data[\"L90_900\"]]\n",
    "\n",
    "Ln = [L_indicators, L4_indicators, L10_indicators, L20_indicators, L30_indicators, L60_indicators, L120_indicators, L180_indicators,\n",
    "     L300_indicators, L450_indicators, L900_indicators ]\n",
    "\n",
    "for indicator in Ln:\n",
    "    for i in range (len(indicator)):\n",
    "        rmse = np.sqrt (mean_squared_error ( indicator[i] , L_indicators[i]))\n",
    "        \n",
    "        if len(indicator[i].name.split('_')) == 1:\n",
    "            m = indicator[i].name.split('_')[0] \n",
    "            f = 2\n",
    "        else:\n",
    "            f = indicator[i].name.split('_')[1]\n",
    "            m = indicator[i].name.split('_')[0]\n",
    "            \n",
    "        rmse_values.append([m,f, rmse])\n",
    "\n",
    "RMSE = pd.DataFrame(rmse_values , columns=[\"Indicator\", \"Frequency\", \"RMSE\"])\n",
    "\n",
    "RMSE[\"N\"] = n_samples(RMSE[\"Frequency\"])\n",
    "RMSE_ = RMSE[RMSE[\"N\" ] > 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "#Samples:  450.0 \n",
      " Indicator:  Leq \n",
      " R2 =  0.999998172790194 \n",
      " RMSE:  0.006781693391525355\n",
      "#Samples:  450.0 \n",
      " Indicator:  L10 \n",
      " R2 =  0.9999994097888427 \n",
      " RMSE:  0.004471649518581388\n",
      "#Samples:  450.0 \n",
      " Indicator:  L90 \n",
      " R2 =  0.9999948607720744 \n",
      " RMSE:  0.007237496665588793\n",
      "#Samples:  225.0 \n",
      " Indicator:  Leq \n",
      " R2 =  0.9829115994206665 \n",
      " RMSE:  0.6558350910520531\n",
      "#Samples:  225.0 \n",
      " Indicator:  L10 \n",
      " R2 =  0.9986405530127881 \n",
      " RMSE:  0.214607644380389\n",
      "#Samples:  225.0 \n",
      " Indicator:  L90 \n",
      " R2 =  0.9989725457244668 \n",
      " RMSE:  0.10233417411086199\n",
      "#Samples:  90.0 \n",
      " Indicator:  Leq \n",
      " R2 =  0.9781419138956403 \n",
      " RMSE:  0.7417371111005141\n",
      "#Samples:  90.0 \n",
      " Indicator:  L10 \n",
      " R2 =  0.9936152615857463 \n",
      " RMSE:  0.46508850277099045\n",
      "#Samples:  90.0 \n",
      " Indicator:  L90 \n",
      " R2 =  0.9960999284344009 \n",
      " RMSE:  0.19937730680671453\n",
      "#Samples:  45.0 \n",
      " Indicator:  Leq \n",
      " R2 =  0.9641121075286885 \n",
      " RMSE:  0.9504251494285615\n",
      "#Samples:  45.0 \n",
      " Indicator:  L10 \n",
      " R2 =  0.9855029784077189 \n",
      " RMSE:  0.700815086226646\n",
      "#Samples:  45.0 \n",
      " Indicator:  L90 \n",
      " R2 =  0.9931455460967673 \n",
      " RMSE:  0.2643174690652701\n",
      "#Samples:  30.0 \n",
      " Indicator:  Leq \n",
      " R2 =  0.9651566408022736 \n",
      " RMSE:  0.9364917402317983\n",
      "#Samples:  30.0 \n",
      " Indicator:  L10 \n",
      " R2 =  0.9831669450949949 \n",
      " RMSE:  0.7551713745729465\n",
      "#Samples:  30.0 \n",
      " Indicator:  L90 \n",
      " R2 =  0.9917691001312574 \n",
      " RMSE:  0.28964303948953435\n",
      "#Samples:  15.0 \n",
      " Indicator:  Leq \n",
      " R2 =  0.9480045660402961 \n",
      " RMSE:  1.1440013735065364\n",
      "#Samples:  15.0 \n",
      " Indicator:  L10 \n",
      " R2 =  0.9681351730737777 \n",
      " RMSE:  1.0390102483650137\n",
      "#Samples:  15.0 \n",
      " Indicator:  L90 \n",
      " R2 =  0.9132818764922949 \n",
      " RMSE:  0.9401439873559074\n",
      "#Samples:  7.5 \n",
      " Indicator:  Leq \n",
      " R2 =  0.9295342985136841 \n",
      " RMSE:  1.3317810534139733\n",
      "#Samples:  7.5 \n",
      " Indicator:  L10 \n",
      " R2 =  0.950850629453296 \n",
      " RMSE:  1.2903960631827562\n",
      "#Samples:  7.5 \n",
      " Indicator:  L90 \n",
      " R2 =  0.9710534297413242 \n",
      " RMSE:  0.5431727748228641\n",
      "#Samples:  5.0 \n",
      " Indicator:  Leq \n",
      " R2 =  0.9127607462403633 \n",
      " RMSE:  1.481835221550722\n",
      "#Samples:  5.0 \n",
      " Indicator:  L10 \n",
      " R2 =  0.9310653430398366 \n",
      " RMSE:  1.5282093717815162\n",
      "#Samples:  5.0 \n",
      " Indicator:  L90 \n",
      " R2 =  0.9547789411173754 \n",
      " RMSE:  0.6789062406167777\n",
      "#Samples:  3.0 \n",
      " Indicator:  Leq \n",
      " R2 =  0.8904694364158763 \n",
      " RMSE:  1.6603957130987625\n",
      "#Samples:  3.0 \n",
      " Indicator:  L10 \n",
      " R2 =  0.9076747524534818 \n",
      " RMSE:  1.7685782904666136\n",
      "#Samples:  3.0 \n",
      " Indicator:  L90 \n",
      " R2 =  0.9140570413474648 \n",
      " RMSE:  0.9359326272534931\n",
      "#Samples:  2.0 \n",
      " Indicator:  Leq \n",
      " R2 =  0.8466989398631164 \n",
      " RMSE:  1.9643393715149469\n",
      "#Samples:  2.0 \n",
      " Indicator:  L10 \n",
      " R2 =  0.8579671851073681 \n",
      " RMSE:  2.1936048437626896\n",
      "#Samples:  2.0 \n",
      " Indicator:  L90 \n",
      " R2 =  0.7613351909702819 \n",
      " RMSE:  1.5596732755149718\n",
      "#Samples:  1.0 \n",
      " Indicator:  Leq \n",
      " R2 =  0.6446545630135592 \n",
      " RMSE:  2.9906756434676636\n",
      "#Samples:  1.0 \n",
      " Indicator:  L10 \n",
      " R2 =  0.668847585216317 \n",
      " RMSE:  3.3494848088867455\n",
      "#Samples:  1.0 \n",
      " Indicator:  L90 \n",
      " R2 =  0.5665389638832696 \n",
      " RMSE:  2.101911708844959\n"
     ]
    }
   ],
   "source": [
    "col_x = [[\"Leq\", \"L10\", \"L90\"], [\"Leq_4\", \"L10_4\", \"L90_4\"], [\"Leq_10\", \"L10_10\", \"L90_10\"],[\"Leq_20\", \"L10_20\", \"L90_20\"],\n",
    "        [\"Leq_30\", \"L10_30\" , \"L90_30\"], [\"Leq_60\", \"L10_60\", \"L90_60\"], [\"Leq_120\", \"L10_120\", \"L90_120\"],\n",
    "        [ \"Leq_180\", \"L10_180\", \"L90_180\"], [\"Leq_300\", \"L10_300\", \"L90_300\"], [\"Leq_450\", \"L10_450\", \"L90_450\"], [\"Leq_900\", \"L10_900\", \"L90_900\"]]\n",
    "\n",
    "col_y = col_x [0]\n",
    "R2_corr = []\n",
    "RMSE_corr = []\n",
    "errors_corr = []\n",
    "\n",
    "train = data.loc[data[\"week\"] != 8].reset_index()\n",
    "test = data.loc[data[\"week\"] == 8].reset_index()\n",
    "\n",
    "for x in col_x: \n",
    "    train_X = pd.DataFrame(data = train, columns = x)\n",
    "    test_X = pd.DataFrame(data = test , columns = x)\n",
    "                \n",
    "    if len(x[0].split('_')) == 1:\n",
    "        f = 2\n",
    "    else:\n",
    "        f = x[0].split('_')[1]\n",
    "        \n",
    "    n = 900/ int(f)\n",
    "    train_X [\"n\"] = n\n",
    "    test_X [\"n\"] = n\n",
    "    \n",
    "    for y in col_y:\n",
    "        \n",
    "        train_y = train[y]\n",
    "        test_y = test[y]\n",
    "        rf = RandomForestRegressor(n_estimators=100, max_features = 4, min_samples_leaf= 1, n_jobs= 8)\n",
    "        rf.fit(train_X, train_y)\n",
    "        y_pred_reg = rf.predict(test_X)\n",
    "        r2 = r2_score(test_y , y_pred_reg)\n",
    "        rmse = np.sqrt (mean_squared_error(test_y, y_pred_reg))\n",
    "                   \n",
    "        print (\"#Samples: \", n , \"\\n\", \"Indicator: \", y, \"\\n\", \"R2 = \", r2, \"\\n\", \"RMSE: \", rmse )\n",
    "        \n",
    "        R2_corr.append([r2, y , n])\n",
    "        RMSE_corr.append([rmse, y , n])\n",
    "        \n",
    "        for i in range(len(y_pred_reg)):\n",
    "            error = test_y[i] - y_pred_reg[i]\n",
    "            errors_corr.append([error, y , n])\n",
    "            "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'RMSE_corr' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-4-9b34fe1d28ae>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mRMSE_corr\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mDataFrame\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mRMSE_corr\u001b[0m \u001b[0;34m,\u001b[0m \u001b[0mcolumns\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"RMSE_corr\"\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\"Indicator_corr\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"n\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      2\u001b[0m \u001b[0mRMSE_corr_\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mRMSE_corr\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mRMSE_corr\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"n\"\u001b[0m \u001b[0;34m]\u001b[0m \u001b[0;34m>\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0mRMSE_merge\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mconcat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mRMSE\u001b[0m \u001b[0;34m,\u001b[0m \u001b[0mRMSE_corr\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mjoin\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"inner\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"n\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"Indicator\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"RMSE\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"RMSE_corr\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0mRMSE_merge_\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mRMSE_merge\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mRMSE_merge\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"n\"\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m>\u001b[0m \u001b[0;36m4\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'RMSE_corr' is not defined"
     ]
    }
   ],
   "source": [
    "RMSE_corr = pd.DataFrame(RMSE_corr , columns=[\"RMSE_corr\",\"Indicator_corr\", \"n\"])\n",
    "RMSE_corr_ = RMSE_corr[RMSE_corr[\"n\" ] > 1]\n",
    "\n",
    "RMSE_merge = pd.concat([RMSE , RMSE_corr], axis=1, join=\"inner\")[[\"n\", \"Indicator\", \"RMSE\", \"RMSE_corr\"]]\n",
    "RMSE_merge_ = RMSE_merge[RMSE_merge[\"n\"] > 4]\n",
    "Leq_corr = RMSE_merge_[RMSE_merge_[\"Indicator\"] == \"Leq\"]\n",
    "L10_corr = RMSE_merge_[RMSE_merge_[\"Indicator\"] == \"L10\"]\n",
    "L90_corr = RMSE_merge_[RMSE_merge_[\"Indicator\"] == \"L90\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(450, 5)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#sns.set_context(\"notebook\", font_scale=1.5, rc={\"lines.linewidth\": 2.5})\n",
    "sns.set_style(\"whitegrid\")\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "#plt.rcParams['figure.figsize'] = 16,16\n",
    "\n",
    "fig = plt.gcf()\n",
    "fig.set_size_inches(10, 5)\n",
    "\n",
    "plt.plot( 'n', 'RMSE', data=Leq_corr, marker='o', color='blueviolet', label = \"Leq\")\n",
    "plt.plot( 'n', 'RMSE_corr', data=Leq_corr, marker='o', color='blueviolet',  label = \"$\\hat{L}_{eq}$\" , linestyle='--')\n",
    "plt.plot( 'n', 'RMSE', data=L10_corr, marker='x', color='deeppink', label = \"L10\")\n",
    "plt.plot( 'n', 'RMSE_corr', data=L10_corr, marker='x', color='deeppink', label = \"$\\hat{L}_{10}$\" , linestyle='--')\n",
    "plt.plot( 'n', 'RMSE', data=L90_corr, marker='*', color='lightcoral', label = \"L90\")\n",
    "plt.plot( 'n', 'RMSE_corr', data=L90_corr, marker='*', color='lightcoral', label = \"$\\hat{L}_{90}$\", linestyle='--')\n",
    "plt.legend()\n",
    "plt.xscale('log')\n",
    "plt.xlabel(\"n\")\n",
    "plt.ylabel(\"RMSE [dB]\")\n",
    "ax.set_xticks(RMSE_[\"N\"].unique())\n",
    "ax.get_xaxis().set_major_formatter(tk.ScalarFormatter())\n",
    "plt.xlim((450,5))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Histogram of the error after correction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "data1 = data [data[\"week\"] == 8].reset_index()\n",
    "\n",
    "\n",
    "L_indicators = [data1[\"Leq\"], data1[\"L10\"], data1[\"L90\"]]\n",
    "\n",
    "L4_indicators = [data1[\"Leq_4\"], data1[\"L10_4\"], data1[\"L90_4\"]]\n",
    "L10_indicators = [data1[\"Leq_10\"], data1[\"L10_10\"], data1[\"L90_10\"]]\n",
    "L20_indicators = [data1[\"Leq_20\"], data1[\"L10_20\"], data1[\"L90_20\"]]\n",
    "L30_indicators = [data1[\"Leq_30\"], data1[\"L10_30\"], data1[\"L90_30\"]]\n",
    "L60_indicators = [data1[\"Leq_60\"], data1[\"L10_60\"], data1[\"L90_60\"]]\n",
    "L120_indicators = [data1[\"Leq_120\"], data1[\"L10_120\"], data1[\"L90_120\"]]\n",
    "L180_indicators = [data1[\"Leq_180\"], data1[\"L10_180\"], data1[\"L90_180\"]]\n",
    "L300_indicators = [data1[\"Leq_300\"], data1[\"L10_300\"], data1[\"L90_300\"]]\n",
    "L450_indicators = [data1[\"Leq_450\"], data1[\"L10_450\"], data1[\"L90_450\"]]\n",
    "L900_indicators = [data1[\"Leq_900\"], data1[\"L10_900\"], data1[\"L90_900\"]]\n",
    "\n",
    "Ln = [L_indicators, L4_indicators, L10_indicators, L20_indicators, L30_indicators, L60_indicators, L120_indicators, L180_indicators,\n",
    "     L300_indicators, L450_indicators, L900_indicators ]\n",
    "\n",
    "\n",
    "errors_values = []\n",
    "for indicator in Ln:\n",
    "    for i in range (len(indicator)):\n",
    "        if len(indicator[i].name.split('_')) == 1:\n",
    "            m = indicator[i].name.split('_')[0] \n",
    "            f = 2\n",
    "        else:\n",
    "            f = indicator[i].name.split('_')[1]\n",
    "            m = indicator[i].name.split('_')[0]\n",
    "        n = 900 /  int (f)\n",
    "        for j in range (len (indicator[i])):\n",
    "            error = indicator[i][j] - L_indicators[i][j]\n",
    "            errors_values.append([m, n , error])\n",
    "            \n",
    "\n",
    "errors = pd.DataFrame(errors_values , columns=[\"Indicator\", \"n\", \"Error\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Leq225     0.019765\n",
       "Leq90      0.015602\n",
       "Leq30     -0.049945\n",
       "L10_225   -0.017572\n",
       "L10_90    -0.002340\n",
       "L10_30    -0.062324\n",
       "L90_225    0.003073\n",
       "L90_90    -0.006241\n",
       "L90_30    -0.019128\n",
       "dtype: float64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ERR_corr = pd.DataFrame(errors_corr , columns=[\"Error\",\"Indicator\", \"n\"])\n",
    "E_Ln = pd.DataFrame()\n",
    "E_Ln[\"Leq225\"] = ERR_corr[ERR_corr[\"Indicator\"]==\"Leq\"][ERR_corr[\"n\"] == 225][\"Error\"].values\n",
    "E_Ln [\"Leq90\"] = ERR_corr[ERR_corr[\"Indicator\"]==\"Leq\"][ERR_corr[\"n\"] == 90][\"Error\"].values\n",
    "E_Ln [\"Leq30\"] = ERR_corr[ERR_corr[\"Indicator\"]==\"Leq\"][ERR_corr[\"n\"] == 30][\"Error\"].values\n",
    "\n",
    "E_Ln[\"L10_225\"] = ERR_corr[ERR_corr[\"Indicator\"]==\"L10\"][ERR_corr[\"n\"] == 225][\"Error\"].values\n",
    "E_Ln [\"L10_90\"] = ERR_corr[ERR_corr[\"Indicator\"]==\"L10\"][ERR_corr[\"n\"] == 90][\"Error\"].values\n",
    "E_Ln [\"L10_30\"] = ERR_corr[ERR_corr[\"Indicator\"]==\"L10\"][ERR_corr[\"n\"] == 30][\"Error\"].values\n",
    "\n",
    "E_Ln[\"L90_225\"] = ERR_corr[ERR_corr[\"Indicator\"]==\"L90\"][ERR_corr[\"n\"] == 225][\"Error\"].values\n",
    "E_Ln [\"L90_90\"] = ERR_corr[ERR_corr[\"Indicator\"]==\"L90\"][ERR_corr[\"n\"] == 90][\"Error\"].values\n",
    "E_Ln [\"L90_30\"] = ERR_corr[ERR_corr[\"Indicator\"]==\"L90\"][ERR_corr[\"n\"] == 30][\"Error\"].values\n",
    "\n",
    "E_Ln.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x122d07710>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df1 = errors[errors[\"n\"] > 5 ]\n",
    "df1 = df1[df1[\"n\"] <= 225]\n",
    "\n",
    "df2 =  ERR_corr[ ERR_corr[\"n\"] > 5 ]\n",
    "df2 = df2[df2[\"n\"] <= 225]\n",
    "\n",
    "\n",
    "sns.set(style=\"whitegrid\")\n",
    "sns.set_palette([\"blueviolet\", \"deeppink\", \"orangered\"])\n",
    "fig, ax = plt.subplots(2,1)\n",
    "fig.set_size_inches(20, 10)\n",
    "\n",
    "\n",
    "\n",
    "ax1 = sns.boxplot(x=\"n\", y=\"Error\",  hue=\"Indicator\", data=df1, width = 0.5 , ax=ax[0], linewidth=0.9, fliersize=2)\n",
    "ax1.set_xticklabels([\"n = 7.5\" , \"n = 15\", \"n = 30\" , \"n = 45\" , \"n = 90\" , \"n = 225\"])\n",
    "ax1.invert_xaxis()\n",
    "ax1.set_ylim((-3,3))\n",
    "ax1.set(ylabel = \"$e^{p,450}$[dB]\", xlabel = \"\")\n",
    "\n",
    "handles, _ = ax1.get_legend_handles_labels()\n",
    "ax1.legend(handles, ['Leq', 'L10' , 'L90'], title = \"Indicator\")\n",
    "\n",
    "\n",
    "ax2 = sns.boxplot(x=\"n\", y=\"Error\",  hue=\"Indicator\", data=df2, width = 0.5 , ax=ax[1], linewidth=0.9, fliersize=2)\n",
    "ax2.set_ylim((-3,3))\n",
    "ax2.set( ylabel = \"$e^{p,450}$[dB]\" , xlabel= \"\")\n",
    "ax2.set_xticklabels([\"n = 7.5\" , \"n = 15\", \"n = 30\" , \"n = 45\" , \"n = 90\" , \"n = 225\"])\n",
    "ax2.invert_xaxis()\n",
    "\n",
    "\n",
    "handles, _ = ax2.get_legend_handles_labels()\n",
    "ax2.legend(handles, ['$\\hat Leq$', '$\\hat L10$' , '$\\hat L90$'], title = \"Indicator\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Classification for the bias correction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "dd = defaultdict(list)\n",
    "f1_values = []\n",
    "\n",
    "\n",
    "Ll = [data[\"CQI_indicator\"],data[\"CQI_indicator_4\"], data[\"CQI_indicator_10\"],  data[\"CQI_indicator_20\"], data[\"CQI_indicator_30\"],data[\"CQI_indicator_60\"], data[\"CQI_indicator_120\"], data[\"CQI_indicator_180\"], data[\"CQI_indicator_300\"],\n",
    "     data[\"CQI_indicator_450\"] , data[\"CQI_indicator_900\"] ]\n",
    "\n",
    "for label in Ll:\n",
    "    f1 = f1_score( label , Ll[0], average=\"micro\")\n",
    "    cm = confusion_matrix(label, Ll[0], labels= ['Excellent', 'Good', 'Fair', 'Poor'])\n",
    "    if len(label.name.split('_')) == 2:\n",
    "        m = label.name.split('_')[0] \n",
    "        f = 2\n",
    "    else:\n",
    "        f = label.name.split('_')[2]\n",
    "        m = label.name.split('_')[0]\n",
    "    \n",
    "    n = 900 / int(f)\n",
    "    \n",
    "    f1_values.append([m, f, f1])\n",
    "    dd[n].append(cm)\n",
    "\n",
    "\n",
    "F1 = pd.DataFrame(f1_values , columns=[\"Label\", \"Frequency\", \"F1\"])\n",
    "F1[\"N\"] = n_samples(F1[\"Frequency\"])\n",
    "F1 = F1[F1[\"N\"] > 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "#Samples:  450.0 \n",
      " F1:  1.0\n",
      "#Samples:  225.0 \n",
      " F1:  0.9805680119581465\n",
      "#Samples:  90.0 \n",
      " F1:  0.9805680119581465\n",
      "#Samples:  45.0 \n",
      " F1:  0.9641255605381166\n",
      "#Samples:  30.0 \n",
      " F1:  0.968609865470852\n",
      "#Samples:  15.0 \n",
      " F1:  0.9476831091180867\n",
      "#Samples:  7.5 \n",
      " F1:  0.9237668161434976\n",
      "#Samples:  5.0 \n",
      " F1:  0.9252615844544095\n",
      "#Samples:  3.0 \n",
      " F1:  0.9043348281016442\n",
      "#Samples:  2.0 \n",
      " F1:  0.8849028400597907\n",
      "#Samples:  1.0 \n",
      " F1:  0.8295964125560538\n"
     ]
    }
   ],
   "source": [
    "col_x = [[\"Leq\", \"L10\", \"L90\"],\n",
    "         [\"Leq_4\", \"L10_4\", \"L90_4\"],\n",
    "         [\"Leq_10\", \"L10_10\", \"L90_10\"],\n",
    "         [\"Leq_20\", \"L10_20\", \"L90_20\"],\n",
    "        [\"Leq_30\", \"L10_30\" , \"L90_30\"], \n",
    "         [\"Leq_60\", \"L10_60\", \"L90_60\"],\n",
    "         [\"Leq_120\", \"L10_120\", \"L90_120\"],\n",
    "        [\"Leq_180\", \"L10_180\", \"L90_180\"], \n",
    "         [\"Leq_300\", \"L10_300\", \"L90_300\"], \n",
    "         [\"Leq_450\", \"L10_450\", \"L90_450\"],\n",
    "         [\"Leq_900\", \"L10_900\", \"L90_900\"]]\n",
    "\n",
    "db = defaultdict(list)\n",
    "\n",
    "F1_corr = []\n",
    "\n",
    "train = data[data['week'] != 8]\n",
    "test = data[data['week'] == 8]\n",
    "\n",
    "train_y = train[\"CQI_indicator\"]\n",
    "test_y = test[\"CQI_indicator\"]\n",
    "\n",
    "for x in col_x: \n",
    "            \n",
    "    train_X = pd.DataFrame(data = train, columns = x)\n",
    "    test_X = pd.DataFrame(data = test , columns = x)\n",
    "                \n",
    "    if len(x[0].split('_')) == 1:\n",
    "        f = 2\n",
    "    else:\n",
    "        f = x[0].split('_')[1]\n",
    "                    \n",
    "    n = 900/ int(f)\n",
    "\n",
    "    train_X [\"n\"] = n\n",
    "    test_X [\"n\"] = n\n",
    "        \n",
    "    clf = RandomForestClassifier(n_estimators=100, max_features = 4, min_samples_leaf= 3, n_jobs= 8)\n",
    "    clf.fit(train_X, train_y)\n",
    "    y_pred_clf = clf.predict(test_X)\n",
    "\n",
    "    f1 = f1_score(y_pred_clf , test_y, average=\"micro\")\n",
    "    cm = confusion_matrix(test_y, y_pred_clf, labels= ['Excellent', 'Good', 'Fair', 'Poor'])\n",
    "\n",
    "    print (\"#Samples: \", n , \"\\n\", \"F1: \", f1 )\n",
    "        \n",
    "    F1_corr.append([f1, n])\n",
    "    db[n].append(cm)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "F1_corr = pd.DataFrame(F1_corr , columns=[\"F1\", \"n\"])\n",
    "F1_corr = F1_corr[F1_corr[\"n\"] > 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_palette([\"mediumvioletred\"])\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "#plt.rcParams['figure.figsize'] = 16,16\n",
    "fig = plt.gcf()\n",
    "fig.set_size_inches(10, 4)\n",
    "\n",
    "plt.plot( 'N', 'F1', data=F1, marker='o', label = \"F1 of $SC^{p,n}$\")\n",
    "plt.plot( \"n\", \"F1\", data=F1_corr, marker = 'o', label = \"F1 of $SC^{p,450}$ with bias compensation\", linestyle='--')\n",
    "\n",
    "plt.xscale('log')\n",
    "plt.xlabel(\"n\")\n",
    "plt.ylabel (\"F1 Score\")\n",
    "plt.xlim(450,5)\n",
    "plt.legend()\n",
    "ax.set_xticks([450, 225, 90, 45, 30, 15, 7.5, 5]) \n",
    "ax.get_xaxis().set_major_formatter(tk.ScalarFormatter())"
   ]
  },
  {
   "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": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}