Assignment_3.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Assignment 3 - CS3920"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_wine\n",
    "from sklearn.model_selection import train_test_split, cross_val_score, GridSearchCV, KFold\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.pipeline import Pipeline, make_pipeline\n",
    "from sklearn.preprocessing import StandardScaler, MinMaxScaler, Normalizer, RobustScaler\n",
    "from sklearn.base import BaseEstimator\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "\n",
    "import numpy as np\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Global constants"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "CV: int = 3\n",
    "RANDOM_STATE: int = 3103"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Useful Functions"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Convert Dictionary to String"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Takes a dictionary and creates a nice string representation \n",
    "- Each key and value are printed with consistent spacing between them"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "def dictionary_to_string(dictionary: dict) -> str:\n",
    "\t\"\"\"Creates a formatted string from a dictionary.\n",
    "\n",
    "\tArgs:\n",
    "\t\tdictionary (dict): dictionary to be formatted\n",
    "\n",
    "\tReturns:\n",
    "\t\tstr: formatted string\n",
    "\t\"\"\"\t\n",
    "\t# Get the keys and values from the dictionary\n",
    "\tkeys: list[str] = list(dictionary.keys())\n",
    "\tvalues: list[float, dict] = list(dictionary.values())\n",
    "\n",
    "\t# Get the column widths\n",
    "\tcolumn_widths: list[int] = [\n",
    "\t\tmax([len(str(key)), len(str(value))])\n",
    "\t\tfor key, value in zip(keys, values)\n",
    "\t]\n",
    "\n",
    "\t# Create the formatted rows\n",
    "\tformatted_rows: list[str] = []\n",
    "\tfor key, value in zip(keys, values):\n",
    "\t\t# Create a list of formatted column values\n",
    "\t\tformatted_columns: list[str] = [\n",
    "\t\t\tf\"{key:{column_widths[0]}}\\t {value:{column_widths[1]}}\"\n",
    "\t\t]\n",
    "\t\t# Join the formatted columns into a single string\n",
    "\t\tformatted_row: str = \" | \".join(formatted_columns)\n",
    "\t\tformatted_rows.append(formatted_row)\n",
    "\n",
    "\t# Join the formatted rows into a single string\n",
    "\tformatted_table = \"\\n\".join(formatted_rows)\n",
    "\treturn formatted_table\t\t"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Loading Datasets"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Question 1**"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading Wine Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [],
   "source": [
    "wine: np.ndarray = load_wine()\n",
    "X_wine: np.ndarray = wine.data\n",
    "y_wine: np.ndarray = wine.target"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading USPS Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "DATA_SPLIT: int = 10 # percentage of data to be used "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "zip_train: np.ndarray = np.genfromtxt('zip.train', delimiter=' ')\n",
    "zip_train = zip_train[::DATA_SPLIT]\n",
    "\n",
    "zip_test: np.ndarray = np.genfromtxt('zip.test', delimiter=' ')\n",
    "zip_test = zip_test[::DATA_SPLIT]\n",
    "\n",
    "zip_data: np.ndarray = np.concatenate((zip_train, zip_test), axis=0)\n",
    "zip_data = zip_data[::DATA_SPLIT]\n",
    "\n",
    "X_zip: np.ndarray = zip_data[:, 1:]\n",
    "y_zip: np.ndarray = zip_data[:, 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "del zip_train, zip_test"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Splitting Datasets "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Question 2**"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- This functions splits the dataset into training set and test set for each matrix of features `X` and labels  `y`\n",
    "- It takes the features `X` and labels `y` and returns the split data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "def split(X: np.ndarray, y: np.ndarray) -> tuple[np.ndarray[float]]:\n",
    "\t\"\"\"Splits the dataset into training and testing sets.\n",
    "\n",
    "\tArgs:\n",
    "\t\tX (np.ndarray): matrix of features\n",
    "\t\ty (np.ndarray): labels \n",
    "\n",
    "\tReturns:\n",
    "\t\ttuple[np.ndarray[float]]: X and y training and testing sets (4 in total)\n",
    "\t\"\"\"\t\n",
    "\treturn train_test_split(X, y, test_size=0.3, random_state=RANDOM_STATE)\n",
    "\t"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Splitting Wine Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_wine_train, X_wine_test, y_wine_train, y_wine_test = split(X_wine, y_wine)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Splitting USPS Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_zip_train, X_zip_test, y_zip_train, y_zip_test = split(X_zip, y_zip)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Cross Validation"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Performs cross-validation on a dataset for some defined model\n",
    "- All the test and training data is provided along with the instance of the desired model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Cross_Validation:\n",
    "\t\"\"\"Class for performing cross validation on a dataset.\"\"\"\n",
    "\tdef __init__ (self, X_train: np.ndarray, y_train: np.ndarray, X_test: np.ndarray, y_test: np.ndarray, model: BaseEstimator):\n",
    "\t\tself.X_train = X_train\n",
    "\t\tself.y_train = y_train\n",
    "\t\tself.X_test = X_test\n",
    "\t\tself.y_test = y_test\n",
    "\t\tself.model: BaseEstimator = model\n",
    "\t\tself.model.fit(self.X_train, self.y_train)\n",
    "\n",
    "\tdef scores_list(self) -> float:\n",
    "\t\t\"\"\"Computes a list of the scores of the model.\n",
    "\n",
    "\t\tReturns:\n",
    "\t\t\tfloat: list of the scores of the model\n",
    "\t\t\"\"\"\t\t\n",
    "\t\treturn cross_val_score(self.model, self.X_train, self.y_train, cv=CV)\n",
    "\n",
    "\t@property\n",
    "\tdef score(self) -> float:\n",
    "\t\t\"\"\"Computes the score of the model.\n",
    "\n",
    "\t\tReturns:\n",
    "\t\t\tfloat: score of the model\n",
    "\t\t\"\"\"\t\t\n",
    "\t\tscores: np.ndarray = cross_val_score(self.model, self.X_train, self.y_train, cv=CV)\n",
    "\t\treturn (self.scores_list()).mean()\n",
    "\n",
    "\t@property\n",
    "\tdef error_rate(self) -> float:\n",
    "\t\t\"\"\"Computes the error rate of the model.\n",
    "\n",
    "\t\tReturns:\n",
    "\t\t\tfloat: error rate of the model\n",
    "\t\t\"\"\"\t\t\n",
    "\t\treturn (1 - self.score)\n",
    "\n",
    "\t@property\n",
    "\tdef test_score(self) -> float:\n",
    "\t\t\"\"\"Computes the accuracy of the model against the test set.\n",
    "\n",
    "\t\tReturns:\n",
    "\t\t\tfloat: accuracy of the model against the test set\n",
    "\t\t\"\"\"\t\t\n",
    "\t\treturn self.model.score(self.X_test, self.y_test)\n",
    "\n",
    "\t@property\n",
    "\tdef test_error_rate(self) -> float:\n",
    "\t\t\"\"\"Computes the error rate of the model against the test set.\n",
    "\n",
    "\t\tReturns:\n",
    "\t\t\tfloat: error rate of the model against the test set\n",
    "\t\t\"\"\"\t\t\n",
    "\t\treturn (1 - self.test_score)\n",
    "\n",
    "\tdef __str__(self) -> str:\n",
    "\t\t\"\"\"Returns a formatted string of the model's scores.\n",
    "\t\t- Score\n",
    "\t\t- Error Rate\n",
    "\t\t- Test Score\n",
    "\t\t- Test Error Rate\n",
    "\n",
    "\t\tReturns:\n",
    "\t\t\tstr: string of the model's scores\n",
    "\t\t\"\"\"\t\t\n",
    "\t\tdata: dict = {\n",
    "\t\t\t\"Score\": self.score,\n",
    "\t\t\t\"Error Rate\": self.error_rate,\n",
    "\t\t\t\"Test Score\": self.test_score,\n",
    "\t\t\t\"Test Error Rate\": self.test_error_rate\n",
    "\t\t}\n",
    "\t\treturn dictionary_to_string(data)\t"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Plots bar graphs for easier comparisons of data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_bar_chart(data: dict, title: str, xlabel: str, ylabel: str, lower_bound: float, upper_bound: float) -> None:\n",
    "\t\"\"\"Plots a bar chart of the scores.\n",
    "\n",
    "\tArgs:\n",
    "\t\tdata (dict): dictionary of the scores\n",
    "\t\ttitle (str): title of the bar chart\n",
    "\t\txlabel (str): x-axis label\n",
    "\t\tylabel (str): y-axis label\n",
    "\t\tfilename (str): filename of the bar chart\n",
    "\t\"\"\"\t\n",
    "\t# Get the keys and values from the dictionary\n",
    "\tkeys: list[str] = list(data.keys())\n",
    "\tvalues: list[float] = list(data.values())\n",
    "\n",
    "\tfig, ax = plt.subplots()\n",
    "\tax.bar(keys, values)\n",
    "\tax.set_title(title)\n",
    "\tax.set_xlabel(xlabel)\n",
    "\tax.set_ylabel(ylabel)\n",
    "\tax.set_ylim(lower_bound, upper_bound)\n",
    "\tfig.tight_layout()\n",
    "\tplt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Question 3 and 4**\n",
    "\n",
    "3. `Score` = Estimation of accuracy parameter\n",
    "4. `Test Error Rate` = Test error rate of the SVM"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cross Validate Wine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score             \t 0.6533101045296168\n",
      "Error Rate        \t 0.3466898954703832\n",
      "Test Score        \t 0.7222222222222222\n",
      "Test Error Rate   \t 0.2777777777777778\n"
     ]
    }
   ],
   "source": [
    "wine_svm_cross_validation: Cross_Validation = Cross_Validation(X_wine_train, y_wine_train, X_wine_test, y_wine_test, SVC())\n",
    "print(wine_svm_cross_validation)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- It appears like the the score is actually lower than the test score\n",
    "  - This means that the error rate of the test is lower \n",
    "- This means that the model is more accurate than expected\n",
    "- Overall, the model is not good as the scores are very low and the error rates are high"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "wine_svm_scores: dict = {\n",
    "\t\"Score\": wine_svm_cross_validation.score,\n",
    "\t\"Test Score\": wine_svm_cross_validation.test_score\n",
    "}\n",
    "plot_bar_chart(wine_svm_scores, \"Wine SVM Scores\", \"Score\", \"Value\", 0.6, 0.8)\n",
    "\n",
    "wine_svm_error_rates: dict = {\n",
    "\t\"Error Rate\": wine_svm_cross_validation.error_rate,\n",
    "\t\"Test Error Rate\": wine_svm_cross_validation.test_error_rate\n",
    "}\n",
    "\n",
    "plot_bar_chart(wine_svm_error_rates, \"Wine SVM Error Rates\", \"Error Rate\", \"Value\", 0.2, 0.4)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cross Validate USPS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score            \t 0.660894660894661\n",
      "Error Rate       \t 0.339105339105339\n",
      "Test Score       \t 0.5517241379310345\n",
      "Test Error Rate  \t 0.4482758620689655\n"
     ]
    }
   ],
   "source": [
    "zip_svm_cross_validation: Cross_Validation = Cross_Validation(X_zip_train, y_zip_train, X_zip_test, y_zip_test, SVC())\n",
    "print(zip_svm_cross_validation)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- It appears like the score of the model is higher than the test score \n",
    "  - This means that the error rate on the training set is lower than the error rate on the test set\n",
    "- This means the model performs worse than expected\n",
    "- Overall, the model is not good as the scores are very low and the error rates are high"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "zip_svm_scores: dict = {\n",
    "\t\"Score\": zip_svm_cross_validation.score,\n",
    "\t\"Test Score\": zip_svm_cross_validation.test_score\n",
    "}\n",
    "plot_bar_chart(zip_svm_scores, \"Zip SVM Scores\", \"Score\", \"Value\", 0.5, 0.7)\n",
    "\n",
    "zip_svm_error_rates: dict = {\n",
    "\t\"Error Rate\": zip_svm_cross_validation.error_rate,\n",
    "\t\"Test Error Rate\": zip_svm_cross_validation.test_error_rate\n",
    "}\n",
    "plot_bar_chart(zip_svm_error_rates, \"Zip SVM Error Rates\", \"Error Rate\", \"Value\", 0.3, 0.5)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Pipeline"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dictionary of Hyperparameters"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Dictionary containing a grid of hyperparameters to be used in a grid search.\n",
    "- The grid contains two hyperparameters: \n",
    "  - `C` (`svc__C`) - determines the strength of regularization, or the penalty applied to the model for making incorrect predictions\n",
    "    - Larger gamma value means that the model is more non-linear, resulting in a more complex and potentially overfitting model\n",
    "    - Smaller gamma value means that the model is more linear, resulting in a simpler and potentially underfitting model\n",
    "  - `gamma` (`svc__gamma`) - determines the strength of the non-linearity of the model\n",
    "    - Larger gamma value means that the model is more non-linear, resulting in a more complex and potentially overfitting model\n",
    "    - Smaller gamma value means that the model is more linear, resulting in a simpler and potentially underfitting model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "param_grid_svm: dict[str, list[float]] = {\n",
    "\t'svc__C': [0.001, 0.01, 0.1, 1, 10, 100], \n",
    "\t'svc__gamma': [0.001, 0.01, 0.1, 1, 10, 100]\n",
    "}"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Creates a pipeline with Support Vector Correlation (SVC) and some specified scaler (normalizer)\n",
    "- It takes the training and test data\n",
    "- Uses Grid Search and Cross-Validation to tune parameters `C` and `gamma`"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Algorithms"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pipeline of Scalers & Support Vector Correlation with Grid Search"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Uses a pipeline to chain the scaler and the some specified classifier (`SVM` or `MLP`)\n",
    "- There are several methods:\n",
    "  - `best_score` - Computes the score of the predictions made on the test set\n",
    "    - Can be used to compare the performance of different models or to choose the final model to use for making predictions on new data\n",
    "  - `best_params` - Identifies the set of hyperparameters that produce the best performance.\n",
    "    - Can be used to choose the final model to use for making predictions on new data\n",
    "  - `test_score` - Computes the score of the predictions made on the test set\n",
    "    - It applies the trained model to the test data and returns a score that indicates how well the model is able to make predictions on the test data\n",
    "    - Can then be used to choose the final model to use for making predictions on new data\n",
    "  - `error_rate` - Computes the error rate of the predictions made on the test set\n",
    "    - $1-score$\n",
    "    - Hence the points about `test_score` still apply"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Super_Pipeline:\n",
    "\t\"\"\"Creates a model that uses a support vector machine (SVM) classifier with a scaler and performs a grid search. \n",
    "\tUses a pipeline to chain the scaler and the `SVM` classifier.\n",
    "\t\"\"\"\t\n",
    "\tdef __init__ (self, param_grid: dict[str, list[float]], scaler: BaseEstimator, X_train: np.ndarray[float], y_train: np.ndarray[float], X_test: np.ndarray[float], y_test: np.ndarray[float], model: BaseEstimator):\t\n",
    "\t\tself.pipeline: Pipeline = make_pipeline(scaler(), model)\n",
    "\t\tself.grid: GridSearchCV = GridSearchCV(self.pipeline, param_grid, cv=CV)\n",
    "\t\tself.grid.fit(X_train, y_train)\n",
    "\t\tself.X_train: np.ndarray[float] = X_train\n",
    "\t\tself.y_train: np.ndarray[float] = y_train\n",
    "\t\tself.X_test: np.ndarray[float] = X_test\n",
    "\t\tself.y_test: np.ndarray[float] = y_test\n",
    "\t\t\n",
    "\t@property\n",
    "\tdef best_score(self) -> float:\n",
    "\t\t\"\"\"Computes the score that was obtained using the best set of hyperparameters. \n",
    "\n",
    "\t\tReturns:\n",
    "\t\t\tfloat: best score\n",
    "\t\t\"\"\"\t\t\n",
    "\t\treturn self.grid.best_score_\n",
    "\n",
    "\t@property\n",
    "\tdef best_params(self) -> dict[str, float]:\n",
    "\t\t\"\"\"Identifies the set of hyperparameters that produce the best performance. \n",
    "\n",
    "\t\tReturns:\n",
    "\t\t\tdict[str, float]: best parameters\n",
    "\t\t\"\"\"\t\t\n",
    "\t\treturn self.grid.best_params_\n",
    "\n",
    "\t@property\n",
    "\tdef test_score(self) -> float:\n",
    "\t\t\"\"\"Computes the score of the predictions made on the test set.\n",
    "\n",
    "\n",
    "\t\tReturns:\n",
    "\t\t\tfloat: score of the predictions made on the test set\n",
    "\t\t\"\"\"\t\t\n",
    "\t\treturn self.grid.score(self.X_test, self.y_test)\n",
    "\n",
    "\t@property\n",
    "\tdef error_rate(self) -> float:\n",
    "\t\t\"\"\"Computes the error rate of the predictions made on the test set.\n",
    "\n",
    "\t\tReturns:\n",
    "\t\t\tfloat: error rate of the predictions made on the test set\n",
    "\t\t\"\"\"\t\t\n",
    "\t\treturn (1 - self.best_score)\n",
    "\n",
    "\tdef __str__(self) -> str:\n",
    "\t\t\"\"\"Returns a string representation of the object.\n",
    "\t\t- Best Cross-Validation Score\n",
    "\t\t- Best Parameters\n",
    "\t\t- Test Set Score\n",
    "\t\t- Test Set Error Rate\n",
    "\n",
    "\t\tReturns:\n",
    "\t\t\tstr: string representation of the object\n",
    "\t\t\"\"\"\t\n",
    "\t\tdata: dict = {\n",
    "\t\t\t\"Best Cross-Validation Score\": f\"{self.best_score}\",\n",
    "\t\t\t\"Best Parameters\": f\"{self.best_params}\",\n",
    "\t\t\t\"Test Set Score\": f\"{self.test_score}\",\n",
    "\t\t\t\"Test Set Error Rate\": f\"{self.error_rate}\"\n",
    "\t\t}\n",
    "\t\treturn dictionary_to_string(data)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Drawing Graphs for Comparisons"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Takes multiple pipelines which have multiple scalers \n",
    "- Compares:\n",
    "  - Cross-Validation Accuracy\n",
    "  - Test Set Error Rate\n",
    "  - Best Parameter C\n",
    "  - Best Parameter Gamma"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Graph_Comparisons:\n",
    "\tdef __init__ (self, super_pipes: dict[str, Super_Pipeline], best_params: dict[str, float]):\n",
    "\t\tself.super_pipes: dict[str, Super_Pipeline] = super_pipes\n",
    "\t\tself.best_params: dict[str, float] = best_params\n",
    "\n",
    "\tdef cross_validation_score(self, lower_bound: float, upper_bound: float) -> None:\n",
    "\t\t\"\"\"Compares the cross-validation scores of the models.\n",
    "\t\tPrints the scores in a table and displays a bar graph.\n",
    "\n",
    "\t\tArgs:\n",
    "\t\t\tlower_bound (float): lower bound of the y-axis\n",
    "\t\t\tupper_bound (float): upper bound of the y-axis\n",
    "\t\t\"\"\"\t\t\n",
    "\t\tscores: dict = {key: pipe.best_score for key, pipe in self.super_pipes.items()}\n",
    "\t\tprint(dictionary_to_string(scores))\n",
    "\n",
    "\t\tplt.bar(self.super_pipes.keys(), [pipe.best_score for pipe in self.super_pipes.values()])\n",
    "\t\tplt.ylabel(\"Cross-Validation Accuracy\")\n",
    "\t\tplt.xlabel(\"Scaler\")\n",
    "\t\tplt.title(\"Cross-Validation Accuracy by Scaler\")\n",
    "\t\tplt.ylim(lower_bound, upper_bound)\n",
    "\t\tplt.show()\n",
    "\n",
    "\tdef test_score(self, lower_bound: float, upper_bound: float) -> None:\n",
    "\t\t\"\"\"Compares the test set scores of the models.\n",
    "\t\tPrints the scores in a table and displays a bar graph.\n",
    "\n",
    "\t\tArgs:\n",
    "\t\t\tlower_bound (float): lower bound of the y-axis\n",
    "\t\t\tupper_bound (float): upper bound of the y-axis\n",
    "\t\t\"\"\"\t\t\n",
    "\t\terror_rates: dict[str, float] = {key: pipe.error_rate for key, pipe in self.super_pipes.items()}\n",
    "\t\tprint(dictionary_to_string(error_rates))\n",
    "\n",
    "\t\tplt.bar(error_rates.keys(), error_rates.values())\n",
    "\t\tplt.ylabel(\"Test Set Error Rate\")\n",
    "\t\tplt.xlabel(\"Scaler\")\n",
    "\t\tplt.title(\"Test Set Error Rate by Scaler\")\n",
    "\t\tplt.ylim(lower_bound, upper_bound)\n",
    "\t\tplt.show()\n",
    "\n",
    "\tdef best_parameter_C(self) -> None:\n",
    "\t\t\"\"\"Compares the best parameter C of the models.\n",
    "\t\tPrints the best parameter C in a table and displays a bar graph.\n",
    "\t\t\"\"\"\t\t\n",
    "\t\tbest_C: dict[str, float] = {key: pipe.best_params['svc__C'] for key, pipe in self.super_pipes.items()}\n",
    "\t\tprint(dictionary_to_string(best_C))\n",
    "\n",
    "\t\tplt.bar(self.super_pipes.keys(), [pipe.best_params['svc__C'] for pipe in self.super_pipes.values()])\n",
    "\t\tplt.bar(self.super_pipes.keys(), [pipe.best_params['svc__C'] for pipe in self.super_pipes.values()])\n",
    "\t\tplt.ylabel(\"Best Parameter C\")\n",
    "\t\tplt.xlabel(\"Scaler\")\n",
    "\t\tplt.title(\"Best Parameter C\")\n",
    "\t\tplt.show()\n",
    "\n",
    "\tdef best_parameter_gamma(self) -> None:\n",
    "\t\t\"\"\"Compares the best parameter gamma of the models.\n",
    "\t\tPrints the best parameter gamma in a table and displays a bar graph.\n",
    "\t\t\"\"\"\t\t\n",
    "\t\tbest_gamma: dict[str, float] = {key: pipe.best_params['svc__gamma'] for key, pipe in self.super_pipes.items()}\n",
    "\t\tprint(dictionary_to_string(best_gamma))\n",
    "\n",
    "\t\tplt.loglog(self.super_pipes.keys(), [pipe.best_params['svc__gamma'] for pipe in self.super_pipes.values()])\n",
    "\t\tplt.ylabel(\"Best Parameter Gamma\")\n",
    "\t\tplt.xlabel(\"Scaler\")\n",
    "\t\tplt.xscale(\"log\")\n",
    "\t\tplt.title(\"Best Parameter Gamma\")\n",
    "\t\tplt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Question 5 and 6**"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Wine Dataset"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### MinMax Scaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.991869918699187                \n",
      "Best Parameters            \t {'svc__C': 10, 'svc__gamma': 0.1}\n",
      "Test Set Score             \t 0.9814814814814815               \n",
      "Test Set Error Rate        \t 0.008130081300813052             \n"
     ]
    }
   ],
   "source": [
    "wine_svm_super_pipe_minmax: Super_Pipeline = Super_Pipeline(param_grid_svm, MinMaxScaler, X_wine_train, y_wine_train, X_wine_test, y_wine_test, SVC())\n",
    "print(wine_svm_super_pipe_minmax)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Standard Scaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.983933410762679                   \n",
      "Best Parameters            \t {'svc__C': 100, 'svc__gamma': 0.001}\n",
      "Test Set Score             \t 0.9814814814814815                  \n",
      "Test Set Error Rate        \t 0.01606658923732096                 \n"
     ]
    }
   ],
   "source": [
    "wine_svm_super_pipe_standard: Super_Pipeline = Super_Pipeline(param_grid_svm, StandardScaler, X_wine_train, y_wine_train, X_wine_test, y_wine_test, SVC())\n",
    "print(wine_svm_super_pipe_standard)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Normalizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.927022841656988                 \n",
      "Best Parameters            \t {'svc__C': 100, 'svc__gamma': 100}\n",
      "Test Set Score             \t 0.9629629629629629                \n",
      "Test Set Error Rate        \t 0.072977158343012                 \n"
     ]
    }
   ],
   "source": [
    "wine_svm_super_pipe_normalizer: Super_Pipeline = Super_Pipeline(param_grid_svm, Normalizer, X_wine_train, y_wine_train, X_wine_test, y_wine_test, SVC())\n",
    "print(wine_svm_super_pipe_normalizer)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Robust Scaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.983739837398374                 \n",
      "Best Parameters            \t {'svc__C': 10, 'svc__gamma': 0.01}\n",
      "Test Set Score             \t 0.9814814814814815                \n",
      "Test Set Error Rate        \t 0.016260162601625994              \n"
     ]
    }
   ],
   "source": [
    "wine_svm_super_pipe_robust: Super_Pipeline = Super_Pipeline(param_grid_svm, RobustScaler, X_wine_train, y_wine_train, X_wine_test, y_wine_test, SVC())\n",
    "print(wine_svm_super_pipe_robust)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Comparison"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [],
   "source": [
    "wine_super_pipes_svm: dict[str, Super_Pipeline] = {\n",
    "\t\"MinMaxScaler\": wine_svm_super_pipe_minmax,\n",
    "\t\"StandardScaler\": wine_svm_super_pipe_standard,\n",
    "\t\"Normalizer\": wine_svm_super_pipe_normalizer,\n",
    "\t\"RobustScaler\": wine_svm_super_pipe_robust\n",
    "}\n",
    "wine_graph: Graph_Comparisons = Graph_Comparisons(wine_super_pipes_svm, param_grid_svm)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Best Cross-Validation Score"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- It appears like the `MinMax` scaler has the highest accuracy\n",
    "  - Best scaler for this task\n",
    "  - Performs slightly better than the other scalers\n",
    "- It appears like `Normalizer` has the lowest accuracy\n",
    "  - Worst Scaler for this task\n",
    "  - Performs significantly worse than the other scalers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MinMaxScaler     \t 0.991869918699187\n",
      "StandardScaler   \t 0.983933410762679\n",
      "Normalizer       \t 0.927022841656988\n",
      "RobustScaler     \t 0.983739837398374\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "wine_graph.cross_validation_score(0.9, 1.0)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Best Parameters"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- The best parameter for `C` for:\n",
    "  - Both `Standard` Scaler and `Normalizer` is 100\n",
    "  - Both `MinMax` scaler and `Robust` scaler is 10 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MinMaxScaler\t             10\n",
      "StandardScaler\t            100\n",
      "Normalizer  \t            100\n",
      "RobustScaler\t             10\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "wine_graph.best_parameter_C()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- The best parameter for `gamma` for:\n",
    "  - `MinMax` scaler is 0.1\n",
    "  - `Standard` scaler is 0.001\n",
    "  - `Normalizer` is 100\n",
    "  - `Robust` scaler is 0.01"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MinMaxScaler\t            0.1\n",
      "StandardScaler\t          0.001\n",
      "Normalizer  \t            100\n",
      "RobustScaler\t           0.01\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "wine_graph.best_parameter_gamma()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Test Set Score"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- `Normalizer` has the highest test error rate making it the least performing scaler by far\n",
    "- `MinMax` has the lowest error rate making it the best performing scaler \n",
    "- This is consistent with the results from the accuracies in the training set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MinMaxScaler        \t 0.008130081300813052\n",
      "StandardScaler      \t 0.01606658923732096\n",
      "Normalizer          \t   0.072977158343012\n",
      "RobustScaler        \t 0.016260162601625994\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "wine_graph.test_score(0.005, 0.08)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## USPS Dataset"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### MinMax Scaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.7373737373737373                  \n",
      "Best Parameters            \t {'svc__C': 100, 'svc__gamma': 0.001}\n",
      "Test Set Score             \t 0.5862068965517241                  \n",
      "Test Set Error Rate        \t 0.26262626262626265                 \n"
     ]
    }
   ],
   "source": [
    "zip_svm_super_pipe_minmax: Super_Pipeline = Super_Pipeline(param_grid_svm, MinMaxScaler, X_zip_train, y_zip_train, X_zip_test, y_zip_test, SVC())\n",
    "print(zip_svm_super_pipe_minmax)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Standard Scaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.7070707070707071                 \n",
      "Best Parameters            \t {'svc__C': 10, 'svc__gamma': 0.001}\n",
      "Test Set Score             \t 0.5517241379310345                 \n",
      "Test Set Error Rate        \t 0.29292929292929293                \n"
     ]
    }
   ],
   "source": [
    "zip_svm_super_pipe_standard: Super_Pipeline = Super_Pipeline(param_grid_svm, StandardScaler, X_zip_train, y_zip_train, X_zip_test, y_zip_test, SVC())\n",
    "print(zip_svm_super_pipe_standard)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Normalizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.4437229437229437                  \n",
      "Best Parameters            \t {'svc__C': 100, 'svc__gamma': 0.001}\n",
      "Test Set Score             \t 0.5862068965517241                  \n",
      "Test Set Error Rate        \t 0.5562770562770563                  \n"
     ]
    }
   ],
   "source": [
    "zip_svm_super_pipe_robust: Super_Pipeline = Super_Pipeline(param_grid_svm, RobustScaler, X_zip_train, y_zip_train, X_zip_test, y_zip_test, SVC())\n",
    "print(zip_svm_super_pipe_robust)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Robust Scaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.7532467532467533                \n",
      "Best Parameters            \t {'svc__C': 100, 'svc__gamma': 0.1}\n",
      "Test Set Score             \t 0.5862068965517241                \n",
      "Test Set Error Rate        \t 0.24675324675324672               \n"
     ]
    }
   ],
   "source": [
    "zip_svm_super_pipe_normalizer: Super_Pipeline = Super_Pipeline(param_grid_svm, Normalizer, X_zip_train, y_zip_train, X_zip_test, y_zip_test, SVC())\n",
    "print(zip_svm_super_pipe_normalizer)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Comparison"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [],
   "source": [
    "zip_super_pipes: dict[str, Super_Pipeline] = {\n",
    "\t\"MinMaxScaler\": zip_svm_super_pipe_minmax,\n",
    "\t\"StandardScaler\": zip_svm_super_pipe_standard,\n",
    "\t\"Normalizer\": zip_svm_super_pipe_normalizer,\n",
    "\t\"RobustScaler\": zip_svm_super_pipe_robust\n",
    "}\n",
    "\n",
    "zip_graph: Graph_Comparisons = Graph_Comparisons(zip_super_pipes, param_grid_svm)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Best Cross-Validation Score"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- `Normalizer` has the highest accuracy out the of all the scalers \n",
    "  - Not significantly better than the other scalers\n",
    "- `Robust` scaler has the lowest accuracy by far out of all the scaler\n",
    "  - It is significantly worse than the other scalers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MinMaxScaler      \t 0.7373737373737373\n",
      "StandardScaler    \t 0.7070707070707071\n",
      "Normalizer        \t 0.7532467532467533\n",
      "RobustScaler      \t 0.4437229437229437\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "zip_graph.cross_validation_score(0.4, 0.8)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Best Parameters"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Except for `Standard` scaler (10), all the other scalers' optimal `C` is 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MinMaxScaler\t            100\n",
      "StandardScaler\t             10\n",
      "Normalizer  \t            100\n",
      "RobustScaler\t            100\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "zip_graph.best_parameter_C()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MinMaxScaler\t          0.001\n",
      "StandardScaler\t          0.001\n",
      "Normalizer  \t            0.1\n",
      "RobustScaler\t          0.001\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "zip_graph.best_parameter_gamma()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Test Set Score"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Normalizer has the lowest error rate on the test set making it the most accurate\n",
    "  - It is not significantly better than the other scalers\n",
    "- Robust scaler is by far the worst as it has the highest error rate \n",
    "  - It is significantly worse than the other scalers\n",
    "- This is consistent with the results from the accuracies on the training set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MinMaxScaler       \t 0.26262626262626265\n",
      "StandardScaler     \t 0.29292929292929293\n",
      "Normalizer         \t 0.24675324675324672\n",
      "RobustScaler       \t  0.5562770562770563\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "zip_graph.test_score(0.2, 0.6)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Neural Network"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Question 8**"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cross-Validation"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Wine Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [],
   "source": [
    "wine_mlp_cross_validation: Cross_Validation = Cross_Validation(X_wine_train, y_wine_train, X_wine_test, y_wine_test, MLPClassifier(random_state=RANDOM_STATE))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score              \t 0.27409988385598144\n",
      "Error Rate         \t 0.7259001161440186\n",
      "Test Score         \t 0.25925925925925924\n",
      "Test Error Rate    \t 0.7407407407407407\n"
     ]
    }
   ],
   "source": [
    "print(wine_mlp_cross_validation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score             \t 0.6533101045296168\n",
      "Error Rate        \t 0.3466898954703832\n",
      "Test Score        \t 0.7222222222222222\n",
      "Test Error Rate   \t 0.2777777777777778\n"
     ]
    }
   ],
   "source": [
    "print(wine_svm_cross_validation)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- It seems like the `SVM` is much more accurate than the `MLP` as it has a higher score (hence lower error rate) when it comes to the training set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot a graph comparing the cross validation scores of the two models\n",
    "plt.bar([\"SVM\", \"MLP\"], [wine_svm_cross_validation.score, wine_mlp_cross_validation.score])\n",
    "plt.title(\"Cross Validation Score Comparison\")\n",
    "plt.ylabel(\"Cross Validation Score\")\n",
    "plt.xlabel(\"Model\")\n",
    "plt.ylim(0.2, 0.7)\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- `SVM` also seems more accurate than `MLP` in the test set which is consistent with the result from the training set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot a graph comparing the test scores of the two models\n",
    "plt.bar([\"SVM\", \"MLP\"], [wine_svm_cross_validation.test_score, wine_mlp_cross_validation.test_score])\n",
    "plt.title(\"Test Score Comparison\")\n",
    "plt.ylabel(\"Test Score\")\n",
    "plt.xlabel(\"Model\")\n",
    "plt.ylim(0.2, 0.8)\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Zip Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "zip_mlp_cross_validation: Cross_Validation = Cross_Validation(X_zip_train, y_zip_train, X_zip_test, y_zip_test, MLPClassifier(random_state=RANDOM_STATE))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score             \t 0.7056277056277057\n",
      "Error Rate        \t 0.2943722943722943\n",
      "Test Score        \t 0.5517241379310345\n",
      "Test Error Rate   \t 0.4482758620689655\n"
     ]
    }
   ],
   "source": [
    "print(zip_mlp_cross_validation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score            \t 0.660894660894661\n",
      "Error Rate       \t 0.339105339105339\n",
      "Test Score       \t 0.5517241379310345\n",
      "Test Error Rate  \t 0.4482758620689655\n"
     ]
    }
   ],
   "source": [
    "print(zip_svm_cross_validation)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- It appears like `MLP` is more accurate than `SVM` as the score is higher (therefore lower error rate) when it comes to the accuracy on the training set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.bar([\"SVM\", \"MLP\"], [zip_svm_cross_validation.score, zip_mlp_cross_validation.score])\n",
    "plt.title(\"Cross Validation Score Comparison\")\n",
    "plt.ylabel(\"Cross Validation Score\")\n",
    "plt.xlabel(\"Model\")\n",
    "plt.ylim(0.6, 0.75)\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- However, it appears like they are both even when it comes to their performance on the test set which is not consistent with the previous result in the training set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.bar([\"SVM\", \"MLP\"], [zip_svm_cross_validation.test_score, zip_mlp_cross_validation.test_score])\n",
    "plt.title(\"Test Score Comparison\")\n",
    "plt.ylabel(\"Test Score\")\n",
    "plt.xlabel(\"Model\")\n",
    "plt.ylim(0.5, 0.6)\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Comparing Scalers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [],
   "source": [
    "param_grid_mlp: dict[str, list[float]] = {\n",
    "\t'mlpclassifier__solver': ['adam', 'lbfgs', 'sgd'],\n",
    "\t'mlpclassifier__hidden_layer_sizes': [[10],[10,10],[10,10,10]],\n",
    "\t'mlpclassifier__activation': ['tanh', 'relu'],\n",
    "}"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Wine Dataset"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### MinMax Scaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [],
   "source": [
    "wine_mlp_super_pipe_minmax: Super_Pipeline = Super_Pipeline(param_grid_mlp, MinMaxScaler, X_wine_train, y_wine_train, X_wine_test, y_wine_test, MLPClassifier(random_state=RANDOM_STATE))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.9758033294618661                                                                                                \n",
      "Best Parameters            \t {'mlpclassifier__activation': 'tanh', 'mlpclassifier__hidden_layer_sizes': [10], 'mlpclassifier__solver': 'lbfgs'}\n",
      "Test Set Score             \t 0.9814814814814815                                                                                                \n",
      "Test Set Error Rate        \t 0.024196670538133902                                                                                              \n"
     ]
    }
   ],
   "source": [
    "print(wine_mlp_super_pipe_minmax)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Standard Scaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [],
   "source": [
    "wine_mlp_super_pipe_standard: Super_Pipeline = Super_Pipeline(param_grid_mlp, StandardScaler, X_wine_train, y_wine_train, X_wine_test, y_wine_test, MLPClassifier(random_state=RANDOM_STATE))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.991869918699187                                                                                                         \n",
      "Best Parameters            \t {'mlpclassifier__activation': 'tanh', 'mlpclassifier__hidden_layer_sizes': [10, 10, 10], 'mlpclassifier__solver': 'lbfgs'}\n",
      "Test Set Score             \t 0.9814814814814815                                                                                                        \n",
      "Test Set Error Rate        \t 0.008130081300813052                                                                                                      \n"
     ]
    }
   ],
   "source": [
    "print(wine_mlp_super_pipe_standard)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Normalizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [],
   "source": [
    "wine_mlp_super_pipe_normalizer: Super_Pipeline = Super_Pipeline(param_grid_mlp, Normalizer, X_wine_train, y_wine_train, X_wine_test, y_wine_test, MLPClassifier(random_state=RANDOM_STATE))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.800038714672861                                                                                                 \n",
      "Best Parameters            \t {'mlpclassifier__activation': 'tanh', 'mlpclassifier__hidden_layer_sizes': [10], 'mlpclassifier__solver': 'lbfgs'}\n",
      "Test Set Score             \t 0.9814814814814815                                                                                                \n",
      "Test Set Error Rate        \t 0.19996128532713897                                                                                               \n"
     ]
    }
   ],
   "source": [
    "print(wine_mlp_super_pipe_normalizer)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Robust Scaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [],
   "source": [
    "wine_mlp_super_pipe_robust: Super_Pipeline = Super_Pipeline(param_grid_mlp, RobustScaler, X_wine_train, y_wine_train, X_wine_test, y_wine_test, MLPClassifier(random_state=RANDOM_STATE))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.9841269841269842                                                                                                    \n",
      "Best Parameters            \t {'mlpclassifier__activation': 'tanh', 'mlpclassifier__hidden_layer_sizes': [10, 10], 'mlpclassifier__solver': 'lbfgs'}\n",
      "Test Set Score             \t 0.9814814814814815                                                                                                    \n",
      "Test Set Error Rate        \t 0.015873015873015817                                                                                                  \n"
     ]
    }
   ],
   "source": [
    "print(wine_mlp_super_pipe_robust)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Comparison"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [],
   "source": [
    "wine_super_pipes_mlp: dict[str, Super_Pipeline] = {\n",
    "\t\"MinMaxScaler\": wine_mlp_super_pipe_minmax,\n",
    "\t\"StandardScaler\": wine_mlp_super_pipe_standard,\n",
    "\t\"Normalizer\": wine_mlp_super_pipe_normalizer,\n",
    "\t\"RobustScaler\": wine_mlp_super_pipe_robust\n",
    "}\n",
    "wine_graph_mlp: Graph_Comparisons = Graph_Comparisons(wine_super_pipes_mlp, param_grid_mlp)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Best Cross Validation Score"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- `Standard` scaler seems to be the best performing scaler as it has the highest score in the training set\n",
    "  - It is not significantly better than the other scalers\n",
    "- `Normalizer` is the worst performing scaler\n",
    "  - It is significantly worse than the other options\n",
    "  - This is consistent with the previous results from `SVN`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MinMaxScaler      \t 0.9758033294618661\n",
      "StandardScaler    \t 0.991869918699187\n",
      "Normalizer        \t 0.800038714672861\n",
      "RobustScaler      \t 0.9841269841269842\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "wine_graph_mlp.cross_validation_score(0.75, 1)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Test Score"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- `Normalizer` has the highest error rate on the test set meaning that it is the worst performing scaler \n",
    "  - It is significantly worse than the other scalers\n",
    "  - This is consistent with its performance in `SVM`\n",
    "- Standard scaler has the lowest error rate meaning that it is the best performing scaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MinMaxScaler        \t 0.024196670538133902\n",
      "StandardScaler      \t 0.008130081300813052\n",
      "Normalizer          \t  0.19996128532713897\n",
      "RobustScaler        \t 0.015873015873015817\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "wine_graph_mlp.test_score(0, 0.2)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Zip Dataset"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### MinMax Scaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [],
   "source": [
    "zip_mlp_super_pipe_minmax: Super_Pipeline = Super_Pipeline(param_grid_mlp, MinMaxScaler, X_zip_train, y_zip_train, X_zip_test, y_zip_test, MLPClassifier(random_state=RANDOM_STATE))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.7669552669552671                                                                                                \n",
      "Best Parameters            \t {'mlpclassifier__activation': 'tanh', 'mlpclassifier__hidden_layer_sizes': [10], 'mlpclassifier__solver': 'lbfgs'}\n",
      "Test Set Score             \t 0.5862068965517241                                                                                                \n",
      "Test Set Error Rate        \t 0.23304473304473294                                                                                               \n"
     ]
    }
   ],
   "source": [
    "print(zip_mlp_super_pipe_minmax)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Standard Scaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [],
   "source": [
    "zip_mlp_super_pipe_standard: Super_Pipeline = Super_Pipeline(param_grid_mlp, StandardScaler, X_zip_train, y_zip_train, X_zip_test, y_zip_test, MLPClassifier(random_state=RANDOM_STATE))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.7676767676767677                                                                                                \n",
      "Best Parameters            \t {'mlpclassifier__activation': 'relu', 'mlpclassifier__hidden_layer_sizes': [10], 'mlpclassifier__solver': 'lbfgs'}\n",
      "Test Set Score             \t 0.5862068965517241                                                                                                \n",
      "Test Set Error Rate        \t 0.23232323232323226                                                                                               \n"
     ]
    }
   ],
   "source": [
    "print(zip_mlp_super_pipe_standard)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Normalizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [],
   "source": [
    "zip_mlp_super_pipe_normalizer: Super_Pipeline = Super_Pipeline(param_grid_mlp, Normalizer, X_zip_train, y_zip_train, X_zip_test, y_zip_test, MLPClassifier(random_state=RANDOM_STATE))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.797979797979798                                                                                                 \n",
      "Best Parameters            \t {'mlpclassifier__activation': 'tanh', 'mlpclassifier__hidden_layer_sizes': [10], 'mlpclassifier__solver': 'lbfgs'}\n",
      "Test Set Score             \t 0.5517241379310345                                                                                                \n",
      "Test Set Error Rate        \t 0.202020202020202                                                                                                 \n"
     ]
    }
   ],
   "source": [
    "print(zip_mlp_super_pipe_normalizer)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Robust Scaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [],
   "source": [
    "zip_mlp_super_pipe_robust: Super_Pipeline = Super_Pipeline(param_grid_mlp, RobustScaler, X_zip_train, y_zip_train, X_zip_test, y_zip_test, MLPClassifier(random_state=RANDOM_STATE))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Cross-Validation Score\t 0.5699855699855699                                                                                                       \n",
      "Best Parameters            \t {'mlpclassifier__activation': 'tanh', 'mlpclassifier__hidden_layer_sizes': [10, 10, 10], 'mlpclassifier__solver': 'adam'}\n",
      "Test Set Score             \t 0.4827586206896552                                                                                                       \n",
      "Test Set Error Rate        \t 0.43001443001443007                                                                                                      \n"
     ]
    }
   ],
   "source": [
    "print(zip_mlp_super_pipe_robust)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Comparison"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {},
   "outputs": [],
   "source": [
    "zip_super_pipes_mlp: dict[str, Super_Pipeline] = {\n",
    "\t\"MinMaxScaler\": zip_mlp_super_pipe_minmax,\n",
    "\t\"StandardScaler\": zip_mlp_super_pipe_standard,\n",
    "\t\"Normalizer\": zip_mlp_super_pipe_normalizer,\n",
    "\t\"RobustScaler\": zip_mlp_super_pipe_robust\n",
    "}\n",
    "zip_graph_mlp: Graph_Comparisons = Graph_Comparisons(zip_super_pipes_mlp, param_grid_mlp)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Best Cross Validation Score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MinMaxScaler      \t 0.7669552669552671\n",
      "StandardScaler    \t 0.7676767676767677\n",
      "Normalizer        \t  0.797979797979798\n",
      "RobustScaler      \t 0.5699855699855699\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "zip_graph_mlp.cross_validation_score(0.55, 0.85)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Test Set Score"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- `Robust` scaler has the highest error rate hence it is the worst performing scaler \n",
    "  - It is significantly worse than the other scalers\n",
    "  - Consistent with its performance in `SVM`\n",
    "- 'Normalizer` has the lowest error rate hence it is the best performing scaler\n",
    "  - It is significantly better than the other scalers\n",
    "- This is consistent with their performances in the accuracy in the training set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MinMaxScaler       \t 0.23304473304473294\n",
      "StandardScaler     \t 0.23232323232323226\n",
      "Normalizer         \t   0.202020202020202\n",
      "RobustScaler       \t 0.43001443001443007\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "zip_graph_mlp.test_score(0.2, 0.45)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Conformal Prediction"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Use the KFold class to split the training set into 5 or 10 folds.\n",
    "2. For each fold, train a predictive model on the remaining folds and use it to make predictions on the held-out fold.\n",
    "3. Use a GridSearchCV object in combination with the decision_function method to compute conformity scores for each prediction.\n",
    "4. For each prediction, calculate the p-value using the conformity scores. \n",
    "5. Calculate the percentage of errors made on the test set.\n",
    "6. Create a calibration curve by plotting the percentage of errors vs the significance level.\n",
    "7. Compute the average false p-value on the test set."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Wine Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split the dataset into 5 or 10 folds\n",
    "kfold: KFold = KFold(n_splits=5, random_state=RANDOM_STATE, shuffle=True)\n",
    "\n",
    "# Loop over each fold\n",
    "for train_index, test_index in kfold.split(X_wine):\n",
    "    # Split the data into train and test sets for this fold\n",
    "    X_train, X_test = X_wine[train_index], X_wine[test_index]\n",
    "    y_train, y_test = y_wine[train_index], y_wine[test_index]\n",
    "\n",
    "    # Use GridSearchCV to optimize the hyperparameters of the SVM classifier\n",
    "    param_grid: dict = {\n",
    "        'C': [0.001, 0.01, 0.1, 1, 10, 100],\n",
    "        'gamma': [0.001, 0.01, 0.1, 1, 10, 100]\n",
    "    }  \n",
    "    svm: SVC = SVC(probability=True)\n",
    "    grid_search: GridSearchCV = GridSearchCV(svm, param_grid, cv=5)\n",
    "    grid_search.fit(X_train, y_train)\n",
    "\n",
    "    # Use the trained SVM classifier to compute conformity scores for each sample in the test set\n",
    "    conformity_scores: np.ndarray = grid_search.best_estimator_.decision_function(X_test)\n",
    "# Use the trained SVM classifier to compute predicted probabilities for each sample in the test set\n",
    "predicted_probs: np.ndarray = grid_search.best_estimator_.predict_proba(X_test)\n",
    "\n",
    "# Compute conformity scores for each sample in the test set\n",
    "conformity_scores: list[float] = []\n",
    "for index, element in enumerate(X_test):\n",
    "    # Get the predicted probabilities for this sample\n",
    "    probs: np.ndarray = predicted_probs[index]\n",
    "    \n",
    "    # Compute the conformity score for this sample by taking the maximum predicted probability\n",
    "    conformity_score: float = max(probs)\n",
    "    conformity_scores.append(conformity_score)\n",
    "\n",
    "# Compute the false p-values for each sample in the test set\n",
    "false_pvals: list[float] = []\n",
    "for index, element in enumerate(X_test):\n",
    "    # Get the conformity score for this sample\n",
    "    conformity_score = index\n",
    "    # Count the number of conformity scores in the training set that are greater than or equal to the conformity score for this sample\n",
    "    num_greater_equal = sum\n",
    "\n",
    "# Compute the error rates for different significance levels\n",
    "error_rates: list[float] = []\n",
    "significance_levels: list[float] = []\n",
    "for significance_level in np.arange(0, 1, 0.01):\n",
    "    # Count the number of conformity scores in the test set that are less than the significance level\n",
    "    num_less_than_significance_level = sum(conformity_score < significance_level for conformity_score in conformity_scores)\n",
    "\n",
    "    # Compute the error rate as the number of conformity scores less than the significance level divided by the total number of samples in the test set\n",
    "    error_rate = num_less_than_significance_level / len(X_test)\n",
    "    error_rates.append(error_rate)\n",
    "    significance_levels.append(significance_level)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Error rates: 0.22771428571428573\n",
      "Significance levels: 0.49499999999999994\n",
      "Minimum error rate: 0.0\n",
      "Average p-value: 0.7672672280249155\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"Error rates:\", np.mean(error_rates))\n",
    "print(\"Significance levels:\", np.mean(significance_levels))\n",
    "print(\"Minimum error rate:\", min(error_rates))\n",
    "print(\"Average p-value:\", sum(conformity_scores) / len(conformity_scores))\n",
    "\n",
    "# Plot the error rates against the significance levels\n",
    "plt.plot(significance_levels, error_rates)\n",
    "plt.xlabel(\"Significance level\")\n",
    "plt.ylabel(\"Error rate\")\n",
    "plt.show()\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Zip Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {},
   "outputs": [],
   "source": [
    "kfold: KFold = KFold(n_splits=5, random_state=RANDOM_STATE, shuffle=True)\n",
    "\n",
    "for train_index, test_index in kfold.split(X_zip):\n",
    "\tX_train, X_test = X_zip[train_index], X_zip[test_index]\n",
    "\ty_train, y_test = y_zip[train_index], y_zip[test_index]\n",
    "\n",
    "\tparam_grid: dict = {\n",
    "\t\t'C': [0.001, 0.01, 0.1, 1, 10, 100],\n",
    "\t\t'gamma': [0.001, 0.01, 0.1, 1, 10, 100]\n",
    "\t}  \n",
    "\tsvm: SVC = SVC(probability=True)\n",
    "\tgrid_search: GridSearchCV = GridSearchCV(svm, param_grid, cv=5)\n",
    "\tgrid_search.fit(X_train, y_train)\n",
    "\n",
    "\tconformity_scores: np.ndarray = grid_search.best_estimator_.decision_function(X_test)\n",
    "\n",
    "predicted_probs: np.ndarray = grid_search.best_estimator_.predict_proba(X_test)\n",
    "\n",
    "conformity_scores: list[float] = []\n",
    "for index, element in enumerate(X_test):\n",
    "\tprobs: np.ndarray = predicted_probs[index]\n",
    "\tconformity_score: float = max(probs)\n",
    "\tconformity_scores.append(conformity_score)\n",
    "\n",
    "false_pvals: list[float] = []\n",
    "for index, element in enumerate(X_test):\n",
    "\tconformity_score = index\n",
    "\tnum_greater_equal = sum(conformity_score >= conformity_score for conformity_score in conformity_scores)\n",
    "\tfalse_pval = num_greater_equal / len(X_train)\n",
    "\tfalse_pvals.append(false_pval)\n",
    "\n",
    "error_rates: list[float] = []\n",
    "significance_levels: list[float] = []\n",
    "for significance_level in np.arange(0, 1, 0.01):\n",
    "\tnum_less_than_significance_level = sum(conformity_score < significance_level for conformity_score in conformity_scores)\n",
    "\terror_rate = num_less_than_significance_level / len(X_test)\n",
    "\terror_rates.append(error_rate)\n",
    "\tsignificance_levels.append(significance_level)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Error rates: 0.581111111111111\n",
      "Significance levels: 0.49499999999999994\n",
      "Minimum error rate: 0.0\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"Error rates:\", np.mean(error_rates))\n",
    "print(\"Significance levels:\", np.mean(significance_levels))\n",
    "print(\"Minimum error rate:\", min(error_rates))\n",
    "\n",
    "plt.plot(significance_levels, error_rates)\n",
    "plt.xlabel(\"Significance level\")\n",
    "plt.ylabel(\"Error rate\")\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.11.0 ('anaconda-pOKeClin-py3.11')",
   "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": "3.11.0"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "892714c652f493a6d49359491524d2a8c982297775bb9a8664a3f7cea223b623"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}