naive_bayes_kBest.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Naïve Bayes algorithm for classification with kBest"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the part 1 we will sudy 8 different models\n",
    "\n",
    "1. Gaussian Naive Bayes for Binary Classification (class2)\n",
    "2. Gaussian Naive Bayes for Binary Classification (class2) with kBest features\n",
    "3. Gaussian Naive Bayes for Binary Classification (class2) with normalized features\n",
    "4. Gaussian Naive Bayes for Binary Classification (class2) with normalized kBest features\n",
    "5. Gaussian Naive Bayes for Multiclass Classification (class4)\n",
    "6. Gaussian Naive Bayes for Multiclass Classification (class4) with kBest features\n",
    "7. Gaussian Naive Bayes for Multiclass Classification (class4) with normalized features\n",
    "8. Gaussian Naive Bayes for Multiclass Classification (class4) with normalized kBest features\n",
    "\n",
    "In the part 2, after we have found best models for binary and multiclass prediction, we will make prediction for npf_test_hidden and to test set."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Import needed libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {},
   "outputs": [],
   "source": [
    "#%reset\n",
    "import numpy as np\n",
    "from sklearn.metrics import accuracy_score, confusion_matrix\n",
    "import pandas as pd\n",
    "from sklearn.naive_bayes import GaussianNB, ComplementNB, MultinomialNB, BernoulliNB, CategoricalNB\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "import matplotlib.pyplot as plt\n",
    "import os\n",
    "import seaborn as sns\n",
    "from sklearn import preprocessing \n",
    "from sklearn.feature_selection import chi2, SelectKBest\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Define help function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {},
   "outputs": [],
   "source": [
    "def data_cleaner(df, hidden):\n",
    "    data = df.copy()\n",
    "    data.set_index([\"id\"],inplace=True)\n",
    "    data.drop([\"date\",\"partlybad\"],1,inplace=True)\n",
    "    # If hidden=\"N\", then create col class2\n",
    "    if hidden==\"N\":\n",
    "        # Set create col class2\n",
    "        data[\"class2\"] = np.where(data[\"class4\"] == \"nonevent\", \"nonevent\",\"event\")\n",
    "        # Set coll2 to first place\n",
    "        cols = ['class2'] + [col for col in data if col != 'class2']\n",
    "        data = data[cols]\n",
    "    # If hidden==Y, then drop empty col class4\n",
    "    if hidden==\"Y\":\n",
    "        data.drop([\"class4\"],1,inplace=True)\n",
    "    return data\n",
    "\n",
    "def normalize_0to1(df):\n",
    "    scaler = MinMaxScaler()\n",
    "    df_norm = pd.DataFrame(scaler.fit_transform(df), columns=df.columns)\n",
    "    return df_norm\n",
    "\n",
    "def find_n_best_feature(X, y, n):\n",
    "    bestfeatures = SelectKBest(score_func=chi2, k=n)\n",
    "    fit = bestfeatures.fit(X, y)\n",
    "    dfscores = pd.DataFrame(fit.scores_)\n",
    "    dfcolumns = pd.DataFrame(X.columns)\n",
    "    featureScores = pd.concat([dfcolumns,dfscores],axis=1)\n",
    "    featureScores.columns = ['col_name','Score']\n",
    "    featureScores= featureScores.nlargest(n,'Score')\n",
    "    cols = list(featureScores[\"col_name\"])\n",
    "    return cols\n",
    "\n",
    "def split_data(df):\n",
    "    # Define X and y\n",
    "    X = df.drop([\"class2\",\"class4\"], axis=1)\n",
    "    y = df[[\"class2\", \"class4\"]]\n",
    "    # Split df to test 60%, val 20% and test 20%\n",
    "    # Generate TrainVal 80% and test 20% \n",
    "    X_trainval, X_test, y_trainval, y_test = train_test_split(X, y, test_size=0.20, random_state=42, stratify=y)\n",
    "    # Furthermore split train into Train and Val\n",
    "    X_train, X_val, y_train, y_val= train_test_split( X_trainval, y_trainval, test_size=0.25, random_state=42, stratify=y_trainval)\n",
    "    return X_train, X_val, X_trainval, X_test, y_train, y_val, y_trainval, y_test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Import data and print length and categorical distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train set\n",
      "Len: 258\n",
      "nonevent    0.500000\n",
      "II          0.248062\n",
      "Ib          0.189922\n",
      "Ia          0.062016\n",
      "Name: class4, dtype: float64 \n",
      "\n",
      "Test set\n",
      "Len: 86\n",
      "nonevent    0.500000\n",
      "II          0.244186\n",
      "Ib          0.197674\n",
      "Ia          0.058140\n",
      "Name: class4, dtype: float64 \n",
      "\n",
      "Validation set\n",
      "Len: 86\n",
      "nonevent    0.500000\n",
      "II          0.244186\n",
      "Ib          0.197674\n",
      "Ia          0.058140\n",
      "Name: class4, dtype: float64 \n",
      "\n",
      "Train+validation set\n",
      "Len: 344\n",
      "nonevent    0.500000\n",
      "II          0.247093\n",
      "Ib          0.191860\n",
      "Ia          0.061047\n",
      "Name: class4, dtype: float64 \n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Define folder of the raw data\n",
    "os.chdir('C:/Users/saukk/Desktop/ML_Project')\n",
    "\n",
    "# Read data\n",
    "npf_train = pd.read_csv(\"data/npf_train.csv\")\n",
    "npf_test_hidden = pd.read_csv(\"data/npf_test_hidden.csv\")\n",
    "\n",
    "# Clean npf_train and add cols class2 and class4\n",
    "npf_train = data_cleaner(npf_train, \"N\")\n",
    "npf_test_hidden = data_cleaner(npf_test_hidden, \"Y\")\n",
    "\n",
    "# Create normalized npf_train and npf_test_hidden\n",
    "scaler = StandardScaler()\n",
    "scaler.fit(npf_train.iloc[:,2:])\n",
    "\n",
    "npf_train_norm = npf_train.copy()\n",
    "npf_train_norm.iloc[:,2:] = scaler.transform(npf_train_norm.iloc[:,2:])\n",
    "\n",
    "npf_test_hidden_norm = npf_test_hidden.copy()\n",
    "npf_test_hidden_norm.iloc[:,] = scaler.transform(npf_test_hidden_norm.iloc[:,])\n",
    "\n",
    "# Split npf_train to train, val and test\n",
    "X_train, X_val, X_trainval, X_test, y_train, y_val, y_trainval, y_test = split_data(npf_train)\n",
    "\n",
    "# Create 0to1 normalized npf_trainval for bestK search\n",
    "X_trainval_0to1 = normalize_0to1(X_trainval)\n",
    "\n",
    "# Split npf_train_norm to train, val and test\n",
    "X_train_norm, X_val_norm, X_trainval_norm, X_test_norm, y_train, y_val, y_trainval, y_test = split_data(npf_train_norm)\n",
    "\n",
    "# Save all the file to folder cleaned_data\n",
    "#npf_train.to_csv(\"cleaned_data/npf_train.csv\", index=True)\n",
    "#npf_test_hidden.to_csv(\"cleaned_data/npf_test_hidden.csv\", index=True)\n",
    "\n",
    "#npf_train_norm.to_csv(\"cleaned_data/npf_train_norm.csv\", index=True)\n",
    "#npf_test_hidden_norm.to_csv(\"cleaned_data/npf_test_hidden_norm.csv\", index=True)\n",
    "\n",
    "#X_trainval.to_csv(\"cleaned_data/X_trainval.csv\", index=True)\n",
    "#X_trainval_norm.to_csv(\"cleaned_data/X_trainval_norm.csv\", index=True)\n",
    "#y_trainval.to_csv(\"cleaned_data/y_trainval.csv\", index=True)\n",
    "\n",
    "#X_test.to_csv(\"cleaned_data/X_test.csv\", index=True)\n",
    "#X_test_norm.to_csv(\"cleaned_data/X_test_norm.csv\", index=True)\n",
    "#y_test.to_csv(\"cleaned_data/y_test.csv\", index=True)\n",
    "\n",
    "# Print sets distribution\n",
    "names = [\"Train set\", \"Test set\", \"Validation set\", \"Train+validation set\"]\n",
    "sets = [y_train, y_test, y_val, y_trainval]\n",
    "for i in range(len(names)):\n",
    "    df = sets[i]\n",
    "    print(names[i])\n",
    "    print(\"Len:\",len(df))\n",
    "    print(df[\"class4\"].value_counts(normalize=True),\"\\n\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Make correlation matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1440 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "corrmat = npf_train.corr()\n",
    "top_corr_features = corrmat.index\n",
    "plt.figure(figsize=(20,20))\n",
    "g=sns.heatmap(npf_train[top_corr_features].corr(),annot=False,cmap=\"RdYlGn\")\n",
    "#plt.savefig('npf_train_correlationmatrix.pdf')  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 1, study the models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Find best features and print accuracy graph\n",
    "\n",
    "SelectKBest method find correlation between every class and feature. Then it gives these feature correlation scores and sorts result by score. Then we search the number of best features, that gives best training accuracy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {},
   "outputs": [],
   "source": [
    "def bestK_and_graph(X_train, y_train, X_val, y_val, X_test, y_test, title):\n",
    "    len_cols = len(X_train.columns)+1\n",
    "    acc_train = []\n",
    "    acc_test = []\n",
    "    acc_val = []\n",
    "\n",
    "    for i in range(1, len_cols):\n",
    "    \n",
    "        # Train filter with full dataset. Find i number of best col names.\n",
    "        X_train_0to1 = normalize_0to1(X_train)\n",
    "        bestFeatureCols = find_n_best_feature(X_train_0to1, y_train, i)\n",
    "    \n",
    "        # Filter train, val and test set with bestFeaturesFilter\n",
    "        X_train_i = X_train[bestFeatureCols]\n",
    "        X_val_i = X_val[bestFeatureCols]\n",
    "        X_test_i = X_test[bestFeatureCols]\n",
    "        \n",
    "        # Train Gaussian Naive Bayes model\n",
    "        model = GaussianNB()\n",
    "        model.fit(X_train_i, y_train)\n",
    "    \n",
    "        # Save accurancy to list\n",
    "        acc_train.append(accuracy_score(y_train, model.predict(X_train_i)))\n",
    "        acc_test.append(accuracy_score(y_test, model.predict(X_test_i)))\n",
    "        acc_val.append(accuracy_score(y_val, model.predict(X_val_i)))\n",
    "    \n",
    "    # Get max value\n",
    "    max_val = max(acc_val)\n",
    "    \n",
    "    # Loop max values and select largest index with max value\n",
    "    acc_val_max = 0\n",
    "    acc_val_max_index = 0\n",
    "    for i in range(len(acc_val)):\n",
    "        if acc_val[i] >= acc_val_max:\n",
    "            acc_val_max = acc_val[i]\n",
    "            acc_val_max_index = i\n",
    "    bestK_max = acc_val_max_index+1\n",
    "    \n",
    "    # Get the col names that are in bestK\n",
    "    bestCols = find_n_best_feature(X_train_0to1, y_train, bestK_max)\n",
    "    \n",
    "    # Plot the results\n",
    "    plt.figure(num=None, figsize=(15, 10), dpi=80, facecolor='w', edgecolor='k')\n",
    "    x = range(1, len_cols)\n",
    "    axes = plt.gca()\n",
    "    axes.set_ylim([0.56,0.99])\n",
    "    plt.plot(x, acc_train, color=\"black\")\n",
    "    plt.plot(x, acc_test, color=\"blue\")\n",
    "    plt.plot(x, acc_val, color=\"green\")\n",
    "    # Vertical line to points with max values\n",
    "    plt.axhline(0.84, color='blue', linestyle='--')\n",
    "    plt.axvline(bestK_max, color='pink', linestyle='-')\n",
    "    plt.title(title)\n",
    "    plt.xlabel('Number of features')\n",
    "    plt.ylabel('Accurancy')\n",
    "    plt.legend([\"train\", \"test\", \"val\", \"Orig test acc\", \"Max point\"], loc =\"lower right\") \n",
    "    plt.show()\n",
    "    \n",
    "    return bestCols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 1. Gaussian Naive Bayes for Binary Classification (class2)\n",
    "bestCols1 = bestK_and_graph(X_train, y_train[\"class2\"], X_val, y_val[\"class2\"], X_test, y_test[\"class2\"], \"1. Gaussian Naive Bayes for Binary (class2) classification accurancy with n number of bestK features\")\n",
    "\n",
    "# 2. Gaussian Naive Bayes for Binary Classification (class2) with normalized features\n",
    "bestCols2 = bestK_and_graph(X_train_norm, y_train[\"class2\"], X_val_norm, y_val[\"class2\"], X_test_norm, y_test[\"class2\"], \"2. Gaussian Naive Bayes for Binary Classification (class2) accuracy with normalized n number of bestK features\")\n",
    "\n",
    "# 3. Gaussian Naive Bayes for Multiclass Classification (class4)\n",
    "bestCols3 = bestK_and_graph(X_train, y_train[\"class4\"], X_val, y_val[\"class4\"], X_test, y_test[\"class4\"], \"3. Gaussian Naive Bayes for Multiclass Classification (class4) accurancy with n number of bestK features\")\n",
    "\n",
    "# 4. Gaussian Naive Bayes for Multiclass Classification (class4) with normalized feature\n",
    "bestCols4 = bestK_and_graph(X_train_norm, y_train[\"class4\"], X_val_norm, y_val[\"class4\"], X_test_norm, y_test[\"class4\"], \"4. Gaussian Naive Bayes for Multiclass Classification (class4) acuracy with normalized n number of bestK features\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Gaussian Naive Bayes Binary classification (class2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gaussian Naive Bayes Binary accuracy with all features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gaussian Naive Bayes binary classification\n",
      "Train set accurancy: 0.76\n",
      "Test set accurancy: 0.84\n",
      "Validation set accurancy: 0.8\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = GaussianNB()\n",
    "model.fit(X_train, y_train[\"class2\"])\n",
    "\n",
    "print(\"Gaussian Naive Bayes binary classification\")\n",
    "print(\"Train set accurancy:\", round(accuracy_score(y_train[\"class2\"], model.predict(X_train)),2))\n",
    "print(\"Test set accurancy:\", round(accuracy_score(y_test[\"class2\"], model.predict(X_test)),2))\n",
    "print(\"Validation set accurancy:\", round(accuracy_score(y_val[\"class2\"], model.predict(X_val)), 2))\n",
    "\n",
    "# Make confusion matrix\n",
    "ax = plt.axes()\n",
    "y_pred = model.predict(X_test)\n",
    "confusion_matrix = pd.crosstab(y_test[\"class2\"], y_pred, rownames=['Actual'], colnames=['Predicted'])\n",
    "sns.heatmap(confusion_matrix, annot=True, ax=ax)\n",
    "ax.set_title('Test set confusion matrix')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gaussian Naive Bayes Binary accuracy with kBest features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gaussian Naive Bayes binary classification with 60 best features\n",
      "Train set accurancy: 0.81\n",
      "Test set accurancy: 0.87\n",
      "Validation set accurancy: 0.85\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_train_i = X_train[bestCols1]\n",
    "X_val_i = X_val[bestCols1]\n",
    "X_test_i = X_test[bestCols1]\n",
    "\n",
    "model = GaussianNB()\n",
    "model.fit(X_train_i, y_train[\"class2\"])\n",
    "\n",
    "print(\"Gaussian Naive Bayes binary classification with\", len(bestCols1), \"best features\")\n",
    "print(\"Train set accurancy:\", round(accuracy_score(y_train[\"class2\"], model.predict(X_train_i)),2))\n",
    "print(\"Test set accurancy:\", round(accuracy_score(y_test[\"class2\"], model.predict(X_test_i)),2))\n",
    "print(\"Validation set accurancy:\", round(accuracy_score(y_val[\"class2\"], model.predict(X_val_i)), 2))\n",
    "\n",
    "# Make confusion matrix\n",
    "ax = plt.axes()\n",
    "y_pred = model.predict(X_test_i)\n",
    "confusion_matrix = pd.crosstab(y_test[\"class2\"], y_pred, rownames=['Actual'], colnames=['Predicted'])\n",
    "sns.heatmap(confusion_matrix, annot=True, ax=ax)\n",
    "ax.set_title('Test set confusion matrix')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Gaussian Naive Bayes Binary classification (class2) normalized features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gaussian Naive Bayes Binary accuracy with all normalized features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gaussian Naive Bayes binary classification normalized features\n",
      "Train set accurancy: 0.77\n",
      "Test set accurancy: 0.83\n",
      "Validation set accurancy: 0.81\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = GaussianNB()\n",
    "model.fit(X_train_norm, y_train[\"class2\"])\n",
    "\n",
    "print(\"Gaussian Naive Bayes binary classification normalized features\")\n",
    "print(\"Train set accurancy:\", round(accuracy_score(y_train[\"class2\"], model.predict(X_train_norm)),2))\n",
    "print(\"Test set accurancy:\", round(accuracy_score(y_test[\"class2\"], model.predict(X_test_norm)),2))\n",
    "print(\"Validation set accurancy:\", round(accuracy_score(y_val[\"class2\"], model.predict(X_val_norm)), 2))\n",
    "\n",
    "# Make confusion matrix\n",
    "ax = plt.axes()\n",
    "y_pred = model.predict(X_test_norm)\n",
    "confusion_matrix = pd.crosstab(y_test[\"class2\"], y_pred, rownames=['Actual'], colnames=['Predicted'])\n",
    "sns.heatmap(confusion_matrix, annot=True, ax=ax)\n",
    "ax.set_title('Test set confusion matrix')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gaussian Naive Bayes Binary accuracy with normalized kBest features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gaussian Naive Bayes binary classification with 61 best normalized features\n",
      "Train set accurancy: 0.8\n",
      "Test set accurancy: 0.86\n",
      "Validation set accurancy: 0.86\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_train_i = X_train_norm[bestCols2]\n",
    "X_val_i = X_val_norm[bestCols2]\n",
    "X_test_i = X_test_norm[bestCols2]\n",
    "\n",
    "model = GaussianNB()\n",
    "model.fit(X_train_i, y_train[\"class2\"])\n",
    "\n",
    "print(\"Gaussian Naive Bayes binary classification with\", len(bestCols2), \"best normalized features\")\n",
    "print(\"Train set accurancy:\", round(accuracy_score(y_train[\"class2\"], model.predict(X_train_i)),2))\n",
    "print(\"Test set accurancy:\", round(accuracy_score(y_test[\"class2\"], model.predict(X_test_i)),2))\n",
    "print(\"Validation set accurancy:\", round(accuracy_score(y_val[\"class2\"], model.predict(X_val_i)), 2))\n",
    "\n",
    "# Make confusion matrix\n",
    "ax = plt.axes()\n",
    "y_pred = model.predict(X_test_i)\n",
    "confusion_matrix = pd.crosstab(y_test[\"class2\"], y_pred, rownames=['Actual'], colnames=['Predicted'])\n",
    "sns.heatmap(confusion_matrix, annot=True, ax=ax)\n",
    "ax.set_title('Test set confusion matrix')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Gaussian Naive Bayes Multiclass classification (class4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gaussian Naive Bayes Multiclass accuracy with all features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gaussian Naive Bayes Multiclass classification\n",
      "Train set accurancy: 0.52\n",
      "Test set accurancy: 0.53\n",
      "Validation set accurancy: 0.5\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = GaussianNB()\n",
    "model.fit(X_train, y_train[\"class4\"])\n",
    "\n",
    "print(\"Gaussian Naive Bayes Multiclass classification\")\n",
    "print(\"Train set accurancy:\", round(accuracy_score(y_train[\"class4\"], model.predict(X_train)),2))\n",
    "print(\"Test set accurancy:\", round(accuracy_score(y_test[\"class4\"], model.predict(X_test)),2))\n",
    "print(\"Validation set accurancy:\", round(accuracy_score(y_val[\"class4\"], model.predict(X_val)), 2))\n",
    "\n",
    "\n",
    "# Make confusion matrix\n",
    "ax = plt.axes()\n",
    "y_pred = model.predict(X_test)\n",
    "confusion_matrix = pd.crosstab(y_test[\"class4\"], y_pred, rownames=['Actual'], colnames=['Predicted'])\n",
    "sns.heatmap(confusion_matrix, annot=True, ax=ax)\n",
    "ax.set_title('Test set confusion matrix')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gaussian Naive Bayes Multiclass accuracy with kBest features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gaussian Naive Bayes Multiclass classification with 43 best features\n",
      "Train set accurancy: 0.62\n",
      "Test set accurancy: 0.62\n",
      "Validation set accurancy: 0.64\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_train_i = X_train[bestCols3]\n",
    "X_val_i = X_val[bestCols3]\n",
    "X_test_i = X_test[bestCols3]\n",
    "\n",
    "model = GaussianNB()\n",
    "model.fit(X_train_i, y_train[\"class4\"])\n",
    "\n",
    "print(\"Gaussian Naive Bayes Multiclass classification with\", len(bestCols3), \"best features\")\n",
    "print(\"Train set accurancy:\", round(accuracy_score(y_train[\"class4\"], model.predict(X_train_i)),2))\n",
    "print(\"Test set accurancy:\", round(accuracy_score(y_test[\"class4\"], model.predict(X_test_i)),2))\n",
    "print(\"Validation set accurancy:\", round(accuracy_score(y_val[\"class4\"], model.predict(X_val_i)), 2))\n",
    "\n",
    "# Make confusion matrix\n",
    "ax = plt.axes()\n",
    "y_pred = model.predict(X_test_i)\n",
    "confusion_matrix = pd.crosstab(y_test[\"class4\"], y_pred, rownames=['Actual'], colnames=['Predicted'])\n",
    "sns.heatmap(confusion_matrix, annot=True, ax=ax)\n",
    "ax.set_title('Test set confusion matrix')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Gaussian Naive Bayes Multiclass classification (class4) normalized features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gaussian Naive Bayes Multiclass accuracy with all normalized features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gaussian Naive Bayes Multiclass classification normalized features\n",
      "Train set accurancy: 0.52\n",
      "Test set accurancy: 0.53\n",
      "Validation set accurancy: 0.49\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = GaussianNB()\n",
    "model.fit(X_train_norm, y_train[\"class4\"])\n",
    "\n",
    "print(\"Gaussian Naive Bayes Multiclass classification normalized features\")\n",
    "print(\"Train set accurancy:\", round(accuracy_score(y_train[\"class4\"], model.predict(X_train_norm)),2))\n",
    "print(\"Test set accurancy:\", round(accuracy_score(y_test[\"class4\"], model.predict(X_test_norm)),2))\n",
    "print(\"Validation set accurancy:\", round(accuracy_score(y_val[\"class4\"], model.predict(X_val_norm)), 2))\n",
    "\n",
    "# Make confusion matrix\n",
    "ax = plt.axes()\n",
    "y_pred = model.predict(X_test_norm)\n",
    "confusion_matrix = pd.crosstab(y_test[\"class4\"], y_pred, rownames=['Actual'], colnames=['Predicted'])\n",
    "sns.heatmap(confusion_matrix, annot=True, ax=ax)\n",
    "ax.set_title('Test set confusion matrix')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gaussian Naive Bayes Multiclass accuracy with kBest normalized features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gaussian Naive Bayes Multiclass classification with 50 best normalized features\n",
      "Train set accurancy: 0.62\n",
      "Test set accurancy: 0.6\n",
      "Validation set accurancy: 0.65\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_train_i = X_train_norm[bestCols4]\n",
    "X_val_i = X_val_norm[bestCols4]\n",
    "X_test_i = X_test_norm[bestCols4]\n",
    "\n",
    "model = GaussianNB()\n",
    "model.fit(X_train_i, y_train[\"class4\"])\n",
    "\n",
    "print(\"Gaussian Naive Bayes Multiclass classification with\", len(bestCols4), \"best normalized features\")\n",
    "print(\"Train set accurancy:\", round(accuracy_score(y_train[\"class4\"], model.predict(X_train_i)),2))\n",
    "print(\"Test set accurancy:\", round(accuracy_score(y_test[\"class4\"], model.predict(X_test_i)),2))\n",
    "print(\"Validation set accurancy:\", round(accuracy_score(y_val[\"class4\"], model.predict(X_val_i)), 2))\n",
    "\n",
    "# Make confusion matrix\n",
    "ax = plt.axes()\n",
    "y_pred = model.predict(X_test_i)\n",
    "confusion_matrix = pd.crosstab(y_test[\"class4\"], y_pred, rownames=['Actual'], colnames=['Predicted'])\n",
    "sns.heatmap(confusion_matrix, annot=True, ax=ax)\n",
    "ax.set_title('Test set confusion matrix')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 2, make predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train the best binary and multiclass models and make class prediction for npf_test_hidden\n",
    "\n",
    "Next we observe the result for binary prediction and to multiclass prediction. \n",
    "\n",
    "For binary prediction we got best results with not normalized and 60 kBest features. \n",
    "\n",
    "The results were:\n",
    "\n",
    "Gaussian Naive Bayes binary classification with 60 best features <br>\n",
    "Train set accurancy: 0.81 <br>\n",
    "Test set accurancy: 0.87 <br>\n",
    "Validation set accurancy: 0.85 <br>\n",
    "\n",
    "For multiclass prediction we got best results with not normalized and 43 kBest features. \n",
    "\n",
    "The results were:\n",
    "\n",
    "Gaussian Naive Bayes Multiclass classification with 43 best features <br>\n",
    "Train set accurancy: 0.62 <br>\n",
    "Test set accurancy: 0.62 <br> \n",
    "Validation set accurancy: 0.64 <br>\n",
    "\n",
    "Next we will make predicions to npf_test_hidden with these models."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train Gaussian Naive Bayes binary classification model with 60 kBest features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gaussian Naive Bayes binary classification with 60 best features\n",
      "Train+validation set accurancy: 0.82\n",
      "Test set accurancy: 0.87\n"
     ]
    }
   ],
   "source": [
    "X_trainval_i = X_trainval[bestCols1]\n",
    "X_test_i = X_test[bestCols1]\n",
    "npf_test_hidden_i = npf_test_hidden[bestCols1]\n",
    "\n",
    "model = GaussianNB()\n",
    "model.fit(X_trainval_i, y_trainval[\"class2\"])\n",
    "\n",
    "print(\"Gaussian Naive Bayes binary classification with\", len(bestCols1), \"best features\")\n",
    "print(\"Train+validation set accurancy:\", round(accuracy_score(y_trainval[\"class2\"], model.predict(X_trainval_i)),2))\n",
    "print(\"Test set accurancy:\", round(accuracy_score(y_test[\"class2\"], model.predict(X_test_i)),2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Make binary prediction for X_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test set accuracy: 0.872093023255814\n",
      "Should be: 0.87\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>NB_pred_class</th>\n",
       "      <th>NB_p(event)</th>\n",
       "      <th>NB_p(nonevent)</th>\n",
       "      <th>NB_truelabel</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>3.911760e-33</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>nonevent</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>5.927607e-12</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>nonevent</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>event</td>\n",
       "      <td>9.999437e-01</td>\n",
       "      <td>5.629272e-05</td>\n",
       "      <td>event</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>3.048545e-24</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>nonevent</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>event</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.800085e-10</td>\n",
       "      <td>nonevent</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>event</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>3.073092e-11</td>\n",
       "      <td>nonevent</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>event</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.243588e-13</td>\n",
       "      <td>nonevent</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>event</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>6.348066e-16</td>\n",
       "      <td>event</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>84</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>5.530072e-09</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>nonevent</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85</th>\n",
       "      <td>event</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>9.691350e-29</td>\n",
       "      <td>event</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>86 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   NB_pred_class   NB_p(event)  NB_p(nonevent) NB_truelabel\n",
       "0       nonevent  3.911760e-33    1.000000e+00     nonevent\n",
       "1       nonevent  5.927607e-12    1.000000e+00     nonevent\n",
       "2          event  9.999437e-01    5.629272e-05        event\n",
       "3       nonevent  3.048545e-24    1.000000e+00     nonevent\n",
       "4          event  1.000000e+00    1.800085e-10     nonevent\n",
       "..           ...           ...             ...          ...\n",
       "81         event  1.000000e+00    3.073092e-11     nonevent\n",
       "82         event  1.000000e+00    1.243588e-13     nonevent\n",
       "83         event  1.000000e+00    6.348066e-16        event\n",
       "84      nonevent  5.530072e-09    1.000000e+00     nonevent\n",
       "85         event  1.000000e+00    9.691350e-29        event\n",
       "\n",
       "[86 rows x 4 columns]"
      ]
     },
     "execution_count": 182,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get prediction classes\n",
    "test_bin_class = pd.DataFrame(model.predict(X_test_i))\n",
    "\n",
    "# Get prediction probabilities\n",
    "test_bin_prob = pd.DataFrame(model.predict_proba(X_test_i))\n",
    "\n",
    "# Merge classes and probabilities\n",
    "test_bin_pred = pd.merge(test_bin_class, test_bin_prob, left_index=True, right_index=True)\n",
    "\n",
    "# Reset index of true labels to mach prediction index\n",
    "y_bin = y_test.copy()\n",
    "y_bin = y_bin.reset_index()\n",
    "\n",
    "# Add true labels\n",
    "test_bin_pred = pd.merge(test_bin_pred, y_bin[\"class2\"], left_index=True, right_index=True)\n",
    "\n",
    "# Rename columns\n",
    "test_bin_pred.columns = [\"NB_pred_class\", \"NB_p(event)\", \"NB_p(nonevent)\", \"NB_truelabel\"]\n",
    "\n",
    "# Check accuracy\n",
    "ncorrect = test_bin_pred[test_bin_pred[\"NB_pred_class\"] == test_bin_pred[\"NB_truelabel\"]].shape[0]\n",
    "print(\"Test set accuracy:\", ncorrect/y_test.shape[0])\n",
    "print(\"Should be:\", round(accuracy_score(y_test[\"class2\"], model.predict(X_test_i)),2))\n",
    "\n",
    "# Save results\n",
    "test_bin_pred.to_csv(\"prediction_results/NB_test_bin.csv\", index=False)\n",
    "\n",
    "# Show dataframe\n",
    "test_bin_pred"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Make binary prediction for npf_test_hidden"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>NB_pred_class</th>\n",
       "      <th>NB_p(event)</th>\n",
       "      <th>NB_p(nonevent)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>event</td>\n",
       "      <td>9.999998e-01</td>\n",
       "      <td>1.726299e-07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>6.396269e-17</td>\n",
       "      <td>1.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>3.046145e-34</td>\n",
       "      <td>1.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>event</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.131420e-31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>4.063615e-15</td>\n",
       "      <td>1.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>960</th>\n",
       "      <td>event</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>2.160401e-15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>961</th>\n",
       "      <td>event</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.064034e-14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>962</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>1.942802e-21</td>\n",
       "      <td>1.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>963</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>8.338768e-10</td>\n",
       "      <td>1.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>964</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>2.192984e-50</td>\n",
       "      <td>1.000000e+00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>965 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    NB_pred_class   NB_p(event)  NB_p(nonevent)\n",
       "0           event  9.999998e-01    1.726299e-07\n",
       "1        nonevent  6.396269e-17    1.000000e+00\n",
       "2        nonevent  3.046145e-34    1.000000e+00\n",
       "3           event  1.000000e+00    1.131420e-31\n",
       "4        nonevent  4.063615e-15    1.000000e+00\n",
       "..            ...           ...             ...\n",
       "960         event  1.000000e+00    2.160401e-15\n",
       "961         event  1.000000e+00    1.064034e-14\n",
       "962      nonevent  1.942802e-21    1.000000e+00\n",
       "963      nonevent  8.338768e-10    1.000000e+00\n",
       "964      nonevent  2.192984e-50    1.000000e+00\n",
       "\n",
       "[965 rows x 3 columns]"
      ]
     },
     "execution_count": 183,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get prediction classes\n",
    "bin_class = pd.DataFrame(model.predict(npf_test_hidden_i))\n",
    "\n",
    "# Get prediction probabilities\n",
    "bin_prob = pd.DataFrame(model.predict_proba(npf_test_hidden_i))\n",
    "\n",
    "# Merge classes and probabilities\n",
    "NB_bin_pred = pd.merge(bin_class, bin_prob, left_index=True, right_index=True)\n",
    "\n",
    "# Rename columns\n",
    "NB_bin_pred.columns = [\"NB_pred_class\", \"NB_p(event)\", \"NB_p(nonevent)\"]\n",
    "\n",
    "# Save results\n",
    "NB_bin_pred.to_csv(\"prediction_results/NB_bin.csv\", index=False)\n",
    "\n",
    "# Show dataframe\n",
    "NB_bin_pred"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train Gaussian Naive Bayes multiclass classification model with 43 kBest features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gaussian Naive Bayes multiclass classification with 43 best features\n",
      "Train+validation set accurancy: 0.6\n",
      "Test set accurancy: 0.66\n"
     ]
    }
   ],
   "source": [
    "X_trainval_i = X_trainval[bestCols3]\n",
    "X_test_i = X_test[bestCols3]\n",
    "npf_test_hidden_i = npf_test_hidden[bestCols3]\n",
    "\n",
    "model = GaussianNB()\n",
    "model.fit(X_trainval_i, y_trainval[\"class4\"])\n",
    "\n",
    "print(\"Gaussian Naive Bayes multiclass classification with\", len(bestCols3), \"best features\")\n",
    "print(\"Train+validation set accurancy:\", round(accuracy_score(y_trainval[\"class4\"], model.predict(X_trainval_i)),2))\n",
    "print(\"Test set accurancy:\", round(accuracy_score(y_test[\"class4\"], model.predict(X_test_i)),2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Make multiclass prediction for X_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test set accuracy: 0.6627906976744186\n",
      "Should be: 0.66\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>NB_pred_class</th>\n",
       "      <th>NB_p(II)</th>\n",
       "      <th>NB_p(Ia)</th>\n",
       "      <th>NB_p(Ib)</th>\n",
       "      <th>NB_p(nonevent)</th>\n",
       "      <th>NB_truelabel</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>9.137620e-29</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>2.928694e-47</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>nonevent</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>1.362902e-02</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>4.729173e-08</td>\n",
       "      <td>9.863709e-01</td>\n",
       "      <td>nonevent</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>II</td>\n",
       "      <td>9.910437e-01</td>\n",
       "      <td>1.417204e-142</td>\n",
       "      <td>8.704057e-03</td>\n",
       "      <td>2.521999e-04</td>\n",
       "      <td>Ib</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>8.572165e-23</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>2.616634e-37</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>nonevent</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>II</td>\n",
       "      <td>7.152608e-01</td>\n",
       "      <td>4.502350e-15</td>\n",
       "      <td>2.847392e-01</td>\n",
       "      <td>3.003522e-15</td>\n",
       "      <td>nonevent</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>II</td>\n",
       "      <td>9.193140e-01</td>\n",
       "      <td>5.457505e-07</td>\n",
       "      <td>8.068545e-02</td>\n",
       "      <td>8.818600e-11</td>\n",
       "      <td>nonevent</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>II</td>\n",
       "      <td>9.494042e-01</td>\n",
       "      <td>3.926034e-231</td>\n",
       "      <td>5.059579e-02</td>\n",
       "      <td>2.643854e-12</td>\n",
       "      <td>nonevent</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>Ib</td>\n",
       "      <td>1.831036e-05</td>\n",
       "      <td>1.986767e-04</td>\n",
       "      <td>9.997830e-01</td>\n",
       "      <td>7.858143e-17</td>\n",
       "      <td>Ib</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>84</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>8.278050e-10</td>\n",
       "      <td>1.174208e-04</td>\n",
       "      <td>5.313351e-13</td>\n",
       "      <td>9.998826e-01</td>\n",
       "      <td>nonevent</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85</th>\n",
       "      <td>Ib</td>\n",
       "      <td>5.147084e-02</td>\n",
       "      <td>7.233309e-09</td>\n",
       "      <td>9.485292e-01</td>\n",
       "      <td>1.703697e-25</td>\n",
       "      <td>Ib</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>86 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   NB_pred_class      NB_p(II)       NB_p(Ia)      NB_p(Ib)  NB_p(nonevent)  \\\n",
       "0       nonevent  9.137620e-29   0.000000e+00  2.928694e-47    1.000000e+00   \n",
       "1       nonevent  1.362902e-02   0.000000e+00  4.729173e-08    9.863709e-01   \n",
       "2             II  9.910437e-01  1.417204e-142  8.704057e-03    2.521999e-04   \n",
       "3       nonevent  8.572165e-23   0.000000e+00  2.616634e-37    1.000000e+00   \n",
       "4             II  7.152608e-01   4.502350e-15  2.847392e-01    3.003522e-15   \n",
       "..           ...           ...            ...           ...             ...   \n",
       "81            II  9.193140e-01   5.457505e-07  8.068545e-02    8.818600e-11   \n",
       "82            II  9.494042e-01  3.926034e-231  5.059579e-02    2.643854e-12   \n",
       "83            Ib  1.831036e-05   1.986767e-04  9.997830e-01    7.858143e-17   \n",
       "84      nonevent  8.278050e-10   1.174208e-04  5.313351e-13    9.998826e-01   \n",
       "85            Ib  5.147084e-02   7.233309e-09  9.485292e-01    1.703697e-25   \n",
       "\n",
       "   NB_truelabel  \n",
       "0      nonevent  \n",
       "1      nonevent  \n",
       "2            Ib  \n",
       "3      nonevent  \n",
       "4      nonevent  \n",
       "..          ...  \n",
       "81     nonevent  \n",
       "82     nonevent  \n",
       "83           Ib  \n",
       "84     nonevent  \n",
       "85           Ib  \n",
       "\n",
       "[86 rows x 6 columns]"
      ]
     },
     "execution_count": 185,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get prediction classes\n",
    "test_bin_class = pd.DataFrame(model.predict(X_test_i))\n",
    "\n",
    "# Get prediction probabilities\n",
    "test_bin_prob = pd.DataFrame(model.predict_proba(X_test_i))\n",
    "\n",
    "# Merge classes and probabilities\n",
    "test_bin_pred = pd.merge(test_bin_class, test_bin_prob, left_index=True, right_index=True)\n",
    "\n",
    "# Reset index of true labels to mach prediction index\n",
    "y_bin = y_test.copy()\n",
    "y_bin = y_bin.reset_index()\n",
    "\n",
    "# Add true labels\n",
    "test_bin_pred = pd.merge(test_bin_pred, y_bin[\"class4\"], left_index=True, right_index=True)\n",
    "\n",
    "# Rename columns\n",
    "test_bin_pred.columns = [\"NB_pred_class\", \"NB_p(II)\", \"NB_p(Ia)\",\"NB_p(Ib)\",\"NB_p(nonevent)\", \"NB_truelabel\"]\n",
    "\n",
    "# Check accuracy\n",
    "ncorrect = test_bin_pred[test_bin_pred[\"NB_pred_class\"] == test_bin_pred[\"NB_truelabel\"]].shape[0]\n",
    "print(\"Test set accuracy:\", ncorrect/y_test.shape[0])\n",
    "print(\"Should be:\", round(accuracy_score(y_test[\"class4\"], model.predict(X_test_i)),2))\n",
    "\n",
    "# Save results\n",
    "test_bin_pred.to_csv(\"prediction_results/NB_test_multi.csv\", index=False)\n",
    "\n",
    "# Show dataframe\n",
    "test_bin_pred"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Make multiclass prediction for npf_test_hidden"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>NB_pred_class</th>\n",
       "      <th>NB_p(II)</th>\n",
       "      <th>NB_p(Ia)</th>\n",
       "      <th>NB_p(Ib)</th>\n",
       "      <th>NB_p(nonevent)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>II</td>\n",
       "      <td>9.316851e-01</td>\n",
       "      <td>3.958374e-66</td>\n",
       "      <td>6.831488e-02</td>\n",
       "      <td>1.306913e-09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>1.151875e-20</td>\n",
       "      <td>2.325390e-07</td>\n",
       "      <td>1.644030e-31</td>\n",
       "      <td>9.999998e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>8.620204e-33</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>2.903969e-51</td>\n",
       "      <td>1.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Ib</td>\n",
       "      <td>2.665384e-02</td>\n",
       "      <td>3.907124e-04</td>\n",
       "      <td>9.729554e-01</td>\n",
       "      <td>1.458482e-28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>2.133194e-16</td>\n",
       "      <td>8.024377e-15</td>\n",
       "      <td>4.058921e-17</td>\n",
       "      <td>1.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>960</th>\n",
       "      <td>Ib</td>\n",
       "      <td>1.681069e-01</td>\n",
       "      <td>3.687284e-05</td>\n",
       "      <td>8.318563e-01</td>\n",
       "      <td>2.173895e-14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>961</th>\n",
       "      <td>Ib</td>\n",
       "      <td>3.116180e-01</td>\n",
       "      <td>1.684938e-164</td>\n",
       "      <td>6.883820e-01</td>\n",
       "      <td>4.190588e-15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>962</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>9.863550e-21</td>\n",
       "      <td>1.074880e-08</td>\n",
       "      <td>1.288751e-29</td>\n",
       "      <td>1.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>963</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>8.508702e-10</td>\n",
       "      <td>8.503681e-05</td>\n",
       "      <td>3.714659e-12</td>\n",
       "      <td>9.999150e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>964</th>\n",
       "      <td>nonevent</td>\n",
       "      <td>2.633159e-52</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>5.421465e-60</td>\n",
       "      <td>1.000000e+00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>965 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    NB_pred_class      NB_p(II)       NB_p(Ia)      NB_p(Ib)  NB_p(nonevent)\n",
       "0              II  9.316851e-01   3.958374e-66  6.831488e-02    1.306913e-09\n",
       "1        nonevent  1.151875e-20   2.325390e-07  1.644030e-31    9.999998e-01\n",
       "2        nonevent  8.620204e-33   0.000000e+00  2.903969e-51    1.000000e+00\n",
       "3              Ib  2.665384e-02   3.907124e-04  9.729554e-01    1.458482e-28\n",
       "4        nonevent  2.133194e-16   8.024377e-15  4.058921e-17    1.000000e+00\n",
       "..            ...           ...            ...           ...             ...\n",
       "960            Ib  1.681069e-01   3.687284e-05  8.318563e-01    2.173895e-14\n",
       "961            Ib  3.116180e-01  1.684938e-164  6.883820e-01    4.190588e-15\n",
       "962      nonevent  9.863550e-21   1.074880e-08  1.288751e-29    1.000000e+00\n",
       "963      nonevent  8.508702e-10   8.503681e-05  3.714659e-12    9.999150e-01\n",
       "964      nonevent  2.633159e-52   0.000000e+00  5.421465e-60    1.000000e+00\n",
       "\n",
       "[965 rows x 5 columns]"
      ]
     },
     "execution_count": 186,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get prediction classes\n",
    "multi_class = pd.DataFrame(model.predict(npf_test_hidden_i))\n",
    "\n",
    "# Get prediction probabilities\n",
    "multi_prob = pd.DataFrame(model.predict_proba(npf_test_hidden_i))\n",
    "\n",
    "# Merge classes and probabilities\n",
    "NB_multi_pred = pd.merge(multi_class, multi_prob, left_index=True, right_index=True)\n",
    "\n",
    "# Rename columns\n",
    "NB_multi_pred.columns = [\"NB_pred_class\", \"NB_p(II)\", \"NB_p(Ia)\",\"NB_p(Ib)\",\"NB_p(nonevent)\"]\n",
    "\n",
    "# Save results\n",
    "NB_multi_pred.to_csv(\"prediction_results/NB_multi.csv\", index=False)\n",
    "\n",
    "# Show dataframe\n",
    "NB_multi_pred"
   ]
  },
  {
   "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": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}