5c. Model Dev_Total.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "a06c1ec9",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os, json\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.model_selection import train_test_split\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.svm import SVC\n",
    "from sklearn import svm\n",
    "import time\n",
    "from sklearn.metrics import accuracy_score, confusion_matrix, ConfusionMatrixDisplay, classification_report,\\\n",
    "    precision_score, recall_score, roc_auc_score\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.model_selection import  GridSearchCV, cross_val_predict\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "import lightgbm as lgb\n",
    "from xgboost import XGBClassifier\n",
    "import time\n",
    "from collections import defaultdict\n",
    "from sklearn.preprocessing import StandardScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "294b4bb5",
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.set_option('display.max_columns', None)\n",
    "games = pd.read_csv('data/full_data.csv',index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "424ca3b3",
   "metadata": {},
   "outputs": [],
   "source": [
    "games['total_win'] = np.where(games['total_pts']>games['OverUnder'],1,0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "405c6057",
   "metadata": {},
   "outputs": [],
   "source": [
    "away_offense = ['QB1_away','RB1_away','RB2_away', 'WR1_away', 'WR2_away','WR3_away', 'WR4_away', 'TE1_away','TE2_away']\n",
    "home_offense = ['QB1_home','RB1_home', 'RB2_home', 'WR1_home','WR2_home', 'WR3_home', 'WR4_home','TE1_home', 'TE2_home']\n",
    "games['away_off_total'] = games.filter(away_offense).sum(axis = 1)\n",
    "games['home_off_total'] = games.filter(away_offense).sum(axis = 1)\n",
    "games['total_offense_sal'] = games['away_off_total'] + games['home_off_total']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "3af70ab9",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\binga\\anaconda3\\lib\\site-packages\\pandas\\util\\_decorators.py:311: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  return func(*args, **kwargs)\n"
     ]
    }
   ],
   "source": [
    "games_2022 = games.loc[(games['season'] == 2022) & (games['Week'] != 13)]\n",
    "games_dev = games.loc[games['season'] != 2022]\n",
    "games_dev.dropna(subset=['QB1_home'], axis = 0, inplace = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "cea606f0",
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_cols = ['DST1_away', 'DST1_home', 'QB1_away', 'QB1_home', 'RB1_away',\n",
    "       'RB1_home', 'RB2_away', 'RB2_home', 'WR1_away', 'WR1_home', 'WR2_away',\n",
    "       'WR2_home', 'WR3_away', 'WR3_home', 'WR4_away', 'WR4_home', 'TE1_away',\n",
    "       'TE1_home', 'TE2_away', 'TE2_home','away_off_total','home_off_total','total_offense_sal','PointSpread',\n",
    "        'AwayTeamMoneyLine','HomeTeamMoneyLine','OverUnder']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "5e13771b",
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train, x_test, y_train, y_test = train_test_split(games_dev.filter(pred_cols).fillna(0),games_dev['total_win'], test_size=0.25, random_state=44)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "85ed0463",
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "scaler.fit(x_train)\n",
    "x_train_scaled = scaler.transform(x_train)\n",
    "x_test_scaled = scaler.transform(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "6cc3a56c",
   "metadata": {},
   "outputs": [],
   "source": [
    "random_state = 44"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "47b28003",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_and_get_metrics(clf):\n",
    "    preds_pp = cross_val_predict(clf, x_train, y_train, cv=5, method='predict_proba')[:,1]\n",
    "    preds = preds_pp.round()\n",
    "    print(classification_report(y_train, preds))\n",
    "\n",
    "    clf_cv_acc = accuracy_score(y_train, preds)\n",
    "    clf_cv_prec = precision_score(y_train, preds)\n",
    "    clf_cv_recall = recall_score(y_train, preds)\n",
    "    clf_cv_auc = roc_auc_score(y_train, preds)\n",
    "\n",
    "    cm = confusion_matrix(y_train, preds, labels=clf.classes_)\n",
    "    disp = ConfusionMatrixDisplay(cm, display_labels=clf.classes_)\n",
    "    disp.plot()\n",
    "    plt.show()\n",
    "\n",
    "    return preds_pp, preds, clf_cv_acc, clf_cv_prec, clf_cv_recall, clf_cv_auc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "7aea23d5",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_and_get_metrics_scaled(clf):\n",
    "    preds_pp = cross_val_predict(clf, x_train_scaled, y_train, cv=5, method='predict_proba')[:,1]\n",
    "    preds = preds_pp.round()\n",
    "    print(classification_report(y_train, preds))\n",
    "\n",
    "    clf_cv_acc = accuracy_score(y_train, preds)\n",
    "    clf_cv_prec = precision_score(y_train, preds)\n",
    "    clf_cv_recall = recall_score(y_train, preds)\n",
    "    clf_cv_auc = roc_auc_score(y_train, preds)\n",
    "\n",
    "    cm = confusion_matrix(y_train, preds, labels=clf.classes_)\n",
    "    disp = ConfusionMatrixDisplay(cm, display_labels=clf.classes_)\n",
    "    disp.plot()\n",
    "    plt.show()\n",
    "\n",
    "    return preds_pp, preds, clf_cv_acc, clf_cv_prec, clf_cv_recall, clf_cv_auc"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb5d458b",
   "metadata": {},
   "source": [
    "### KNN "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "fdb14b01",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "7.209737539291382\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "KNeighborsClassifier(n_jobs=-1, n_neighbors=30, weights='distance')"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knn = KNeighborsClassifier()\n",
    "start_time = time.time()\n",
    "param_grid = {\n",
    "    'n_neighbors' : [num for num in range(2,34)],\n",
    "    'weights' : ['uniform','distance'],\n",
    "    'p': [1,2],\n",
    "    'n_jobs': [-1]\n",
    "}\n",
    "\n",
    "knn_grid = GridSearchCV(knn, param_grid, scoring='balanced_accuracy', n_jobs=-1)\n",
    "knn_grid.fit(x_train, y_train)\n",
    "best_knn = knn_grid.best_estimator_\n",
    "print(time.time() - start_time)\n",
    "best_knn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "9b9a0889",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.51      0.55      0.53       589\n",
      "           1       0.50      0.46      0.48       570\n",
      "\n",
      "    accuracy                           0.51      1159\n",
      "   macro avg       0.50      0.50      0.50      1159\n",
      "weighted avg       0.51      0.51      0.50      1159\n",
      "\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": [
    "knn_run = KNeighborsClassifier(n_neighbors = 21, weights = 'distance')\n",
    "best_knn = knn_run.fit(x_train,y_train)\n",
    "knn_preds_pp, knn_preds, knn_cv_acc, knn_cv_prec, knn_cv_recall, knn_cv_auc = plot_and_get_metrics(best_knn)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a572dd2",
   "metadata": {},
   "source": [
    "### Naive Bayes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "8b9dd7b8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.50      0.33      0.40       589\n",
      "           1       0.49      0.66      0.56       570\n",
      "\n",
      "    accuracy                           0.49      1159\n",
      "   macro avg       0.50      0.50      0.48      1159\n",
      "weighted avg       0.50      0.49      0.48      1159\n",
      "\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": [
    "run_nb = GaussianNB()\n",
    "run_nb.fit(x_train,y_train)\n",
    "best_nb = run_nb\n",
    "nb_preds_pp, nb_preds, nb_cv_acc, nb_cv_prec, nb_cv_recall, nb_cv_auc = plot_and_get_metrics(run_nb)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0b23afdb",
   "metadata": {},
   "source": [
    "### SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "8eaae5ec",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "21.252392053604126\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "SVC(C=1.5, degree=2, kernel='linear', probability=True, random_state=44)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "svm = SVC(probability = True, random_state = random_state, max_iter = -1)\n",
    "start_time = time.time()\n",
    "param_grid = {\n",
    "    'C' : [1,1.2,1.3,1.5,1.65,1.8,2],\n",
    "    'kernel' : ['poly','rbf','linear'],\n",
    "    'degree': [2,3],\n",
    "}\n",
    "\n",
    "svm_grid = GridSearchCV(svm, param_grid, scoring='balanced_accuracy', n_jobs=-1)\n",
    "svm_grid.fit(x_train_scaled, y_train)\n",
    "best_svm = svm_grid.best_estimator_\n",
    "print(time.time() - start_time)\n",
    "best_svm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "82b4629b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.51      0.85      0.63       589\n",
      "           1       0.47      0.14      0.21       570\n",
      "\n",
      "    accuracy                           0.50      1159\n",
      "   macro avg       0.49      0.49      0.42      1159\n",
      "weighted avg       0.49      0.50      0.43      1159\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAATgAAAEKCAYAAACGzUnMAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAAAbDElEQVR4nO3de7hVdb3v8feHxWWJIAIKouAGyUtmgj3mJcuDmklaqZ00Nd10cqe2vXR2+1hoPWW57fFUZm4tb+WRLDO6eklBxdya2yuIFzQVE0VBkIVyE8G11vf8McbCKa411xgw55pzjvV5Pc941phjjjnGb8HDl98Yvzl+H0UEZmZF1KfWDTAzqxYXODMrLBc4MyssFzgzKywXODMrLBc4MyssFzgzqxlJCyQ9IWmupEfSbcMk3SHpufTn0JL9z5E0X9Izkg7r7vgucGZWawdFxMSI2Dt9PRWYFRE7A7PS10jaHTgO+AAwGfiZpKZyB3aBM7N6cyQwLV2fBhxVsv2GiFgXES8A84F9yh2ob7VauCm2GdYUY8f0q3UzLIdnn9iy1k2wHN6KNayPt7Q5xzjsoC2jZXlbpn1nP75uZkRMLrNLALdLCuDKiLgKGBkRiwEiYrGkEem+OwAPlHz25XRbl+qqwI0d04+HZo6pdTMsh8nj9q11EyyHB9bdttnHWLa8jQdnjs60b79Rz+/WcW8tdVVaxDocEBGL0iJ2h6S/lzlcZ4W57LOmdVXgzKwRBG3RnnXnZSX31t57pIhF6c+lkv5Ecsm5RNKotPc2Clia7v4yUNoDGg0sKndy34Mzs1wCaCcyLeVI2lLS4I514BPAk8BNwJR0tynAjen6TcBxkgZIGgfsDDxU7hzuwZlZbu1k7sGVMxL4kyRIatH1ETFD0sPAdEknAy8BxwBExDxJ04GngFbg9IgoezPQBc7McgmCt7NfonZ9nIh/ABM62d4CHNLFZy4ALsh6Dhc4M8slgLZuLj/rhQucmeXW3f21euECZ2a5BNDWIDOBu8CZWW4VGWLoAS5wZpZLEL4HZ2bFFAFvN0Z9c4Ezs7xEW6dPTdUfFzgzyyWAdvfgzKyo3IMzs0JKvujrAmdmBRTA29EY83S4wJlZLoFoa5CJiFzgzCy39vAlqpkVkO/BmVmBiTbfgzOzIkpm9HWBM7MCihDro2wcad1ojDJsZnWlHWVaspDUJOlRSbekr8+T9Eqadj9X0uEl++ZKtncPzsxySQYZKto3+irwNLBVybaLI+JHpTttlGy/PXCnpF3K5TK4B2dmOSWDDFmWbo8kjQaOAH6e4cS5k+1d4Mwsl45BhixLBj8Bvs5759A8Q9Ljkq6RNDTdtgOwsGSfbpPtXeDMLLe2UKYF2EbSIyXLKR3HkPQpYGlEzN7o8JcD44GJwGLgoo6PdNIUJ9ubWeUE4u3IXDrKJdsfAHwmHURoBraS9KuIOLFjB0lXA7ekL51sb2bV1THIkGUpe5yIcyJidESMJRk8uCsiTpQ0qmS3o0nS7sHJ9mZWbcGGy89q+YGkiSS1dAFwKjjZ3sx6SKWfZIiIu4G70/WTyuznZHszq54I/CyqmRVTMsjQGI9qucCZWW6e8NLMCimQJ7w0s+JyD87MCinJRXWBM7NCcrK9mRVUEhvoUVQzK6AI+RLVzIrLX/Q1s0JK5oPzPTgzKyTHBppZQSVfE3EPzswKyM+imlmhOfjZzAopmS7Jl6hmVlCNcg+uMfqZZlY3ktlE+mRasugk2X6YpDskPZf+HFqyr5Pte9o/77M7Wwxqo08faOobXDbjWVa+3sT3TxvLkpf7M3L0er555QIGb93GyuVNnH/KWJ6dO5BDj13OGd9/pdbN7/WO/tKrTP78a0TAgme24KKzd2LM+LWcecEC+g8I2lrhsm+P5dnHBtW6qXUheVSrqsn2U4FZEXGhpKnp62/UXbK9pMlppZ2fNrSwfvC7+Vx+5zNcNuNZAKZfNoK9PrqK/3ff0+z10VX89rIRAPRvDqac/Spf/nbZtDPrIcNHrufIL77KmZ/5AKdN/iB9mmDSp1s4+ZyF/PqSHTj9iD247uLR/MvUhd0frNeoXA+ui2T7I4Fp6fo04KiS7fWRbC+pCfgp8Elgd+D4tAL3CvfPHMLHj10OwMePXc79M4YA0DywnT32XUP/AWXzaq0HNTVB/+Z2+jQFA5rbaFnaHwIGDko6BlsObqNlSb8at7K+tKNMC2WCn1M/4b3J9iMjYjFA+nNEuj13sn01L1H3AeZHxD8AJN1AUoGfquI5a0PBucePB8ERJ7Vw+IktvL6sH8NHtgIwfGQrb7T4bkA9alnSn99fvR3X3TeXdW/1Yc69Q5hz7xBeW9SfC6Y9w5fPXYj6BF/7XK/5v7lbOUdRuwx+Lk22lzQpw7HqKtm+s2q7bxXPVzMX3/gcw7dr5Y1lfZl63HjGvO+tWjfJMhq0VSv7H/o6XzxwAqtXNvHNn87n4KOWseuENVz5Hzty34xhfOyIFv7twhc456Tdat3culGh2UQ6TbYHlkgaFRGL0xDopen+dZVsn6naSjqlo/v6WkvZDNe6NXy7pKe29TatHDB5BX9/dCBDt3mbliXJ/x8tS/qy9fDWWjbRurDXR1eyZOEAVizvR1trH+6bOYz3f2g1H//sMu6bkQze3fuXYewyYXWNW1o/OjIZsixlj9NFsj1Jgv2UdLcpwI3peu5k+2oWuEzVNiKuioi9I2LvbYc3xuMfpd56sw9vru6zYX32fw1m7G5vsd8nVnLn9GEA3Dl9GPsftqKWzbQuLF3Un932WsOA5jYgmPiRFSx8vpmWpf3Yc99VAEz8yEoWLWiubUPrSACt0SfTsokuBA6V9BxwaPqaiJgHdCTbz6DGyfYPAzunlfYVkgp9QhXPVxOvv9aX7548DoC2Vjjo6Df48EGr2HXCm1xw2lhm3DCcETskXxPp8M/77M6a1X1oXS/unzmE7//mef5pl3U1+g16t2fmDuLe24Zy2S3zaGsVzz81kNt+M4Ln523Jad9+kaa+wfp1fbjk3HG1bmpdqfSElxsl27cAh3SxX65ke0VUbzQvvbb+CdAEXJM2rkt7T2iOh2aOKbeL1ZnJ4wp5W7WwHlh3GyvbWzbrMYRhu42IQ675n5n2/f0BV8zuapChJ1R1aC8ibgVureY5zKxnecJLMyu0RnkW1QXOzHLxhJdmVliBaG1vjHk6XODMLDffgzOzYgpfoppZQfkenJkVmgucmRVSINo8yGBmReVBBjMrpPAgg5kVWbjAmVkxdT/XW71wgTOz3NyDM7NCioC29sYocI0x1mtmdSVHqlaXJDVLekjSY5LmSfpuuv08Sa9Impsuh5d8xsHPZlY9QcUuUdcBB0fEakn9gL9Jui197+KI+FHpznUX/GxmRVSx0JmIiI40n37pUm6K8foJfjaz4orItnRHUpOkuSTRgHdExIPpW2dIelzSNZKGpttyBz+7wJlZbhHKtNBNsn1EtEXERJLUvX0k7QFcDowHJgKLgYvS3esq+NnMCigZRc3cN+oy2f7dx4w3JN0NTC699ybpauCW9GVdBT+bWUFV4hJV0raStk7XtwA+Dvw9TbPvcDTwZLqeO/jZPTgzy61Co6ijgGmSmkg6W9Mj4hZJ10maSHL5uQA4NTlnzJPUEfzcSo2Dn82sgAJVpMBFxOPAXp1sP6nMZ3IFP7vAmVlu1YuLrywXODPLJyAa5FEtFzgzy80P25tZYWX5Em896LLASbqUMpfaEXFWVVpkZnWtgs+iVl25HtwjPdYKM2scATR6gYuIaaWvJW0ZEWuq3yQzq3eNcona7ZMMkvaX9BTwdPp6gqSfVb1lZlanRLRnW2oty6NaPwEOA1oAIuIx4MAqtsnM6l1kXGos0yhqRCyU3lWNyz4eYWYFFsUYZOiwUNJHgJDUHziL9HLVzHqpOuidZZHlEvU04HSSieVeIZmj6fQqtsnM6p4yLrXVbQ8uIpYBX+iBtphZo2ivdQOyyTKKupOkmyW9JmmppBsl7dQTjTOzOtTxPbgsS41luUS9HphOMnfT9sDvgN9Us1FmVt8qlclQbVkKnCLiuohoTZdf0TC3GM2sKhr9ayKShqWrf5U0FbiBpMmfB/7SA20zs3pVB5efWZQbZJhNUtA6fpNTS94L4PxqNcrM6psq0DuT1AzcAwwgqUW/j4jvpJ2r3wJjSaYsPzYiXk8/cw5wMsl3cc+KiJnlzlHuWdRxm/8rmFnhhKAyj2F1lWz/WWBWRFyYXj1OBb6xKcn2mZ5kSLMKdweaO7ZFxC839bcyswZXgR5cRATQWbL9kcCkdPs04G7gG5Qk2wMvSOpItr+/q3Nk+ZrId4BL0+Ug4AfAZ3L/NmZWHBUaZOgi2X5kRCwGSH+OSHevSrL954BDgFcj4n8BE0iumc2st8pe4DYl2b4rVUm2XxsR7ZJaJW1FUmn9RV+z3irfhJe5k+2BJZJGRcTiNAR6abpbVZLtH0nTp68mGVmdQzdp0mZWbIpsS9ljdJFsT5JgPyXdbQpwY7pe+WT7iPjXdPUKSTOArdLAVjPrrSrzJd6uku3vB6ZLOhl4CTgGKpxsL+lD5d6LiDm5fx0zK4RKfA+uTLJ9C8l9/84+U7Fk+4vKtQ04OOtJsmqjnRXtayt9WKuiWLeu1k2wPCr1gGijP8kQEQf1ZEPMrEHUyXOmWTj42czyc4Ezs6JSg0x46QJnZvk1SA8uy6NaknSipG+nr3eUtE/1m2Zm9Sjrd+AqMdK6ubJ80fdnwP7A8enrVcBPq9YiM6t/DTJleZZL1H0j4kOSHgWIiNfT+EAz663qoHeWRZYC93b6TeOA5PEKGiZTx8yqoR4uP7PIUuD+E/gTMELSBSSzi3yrqq0ys/oVBRpFjYhfS5pN8uiEgKMiwsn2Zr1ZUXpwknYE3gRuLt0WES9Vs2FmVseKUuBIErQ6wmeagXHAMyTzoptZL1SYe3AR8cHS1+ksI6d2sbuZWd3I/SRDRMyR9OFqNMbMGkRRenCSvlbysg/wIeC1qrXIzOpbkUZRgcEl660k9+T+UJ3mmFlDKEIPLv2C76CIOLuH2mNmdU5ULNl+DPBLYDuShweuiohLJJ0HfJl3rhTPjYhb089UJtleUt+IaC03dbmZ9VKV6cG1Av+e3tcfDMyWdEf63sUR8aPSnSudbP8Qyf22uZJuAn4HrOl4MyL+uCm/kZk1uArNFJKGOncEPK+S9DTlg5wrn2wPDANaSDIYPgV8Ov1pZr1Ve8YlI0ljSQJoHkw3nSHpcUnXSBqabqtosv2IdAT1SeCJ9Oe89OeT2ZtuZkWTYz64ssn2AJIGkQxc/u+IWAlcDowHJpL08DoCsCqabN8EDNqUg5pZwWWvAGWT7SX1Iyluv+647RURS0revxq4JX2ZO9m+XIFbHBHfK992M+t1KpSqJUnAL4CnI+LHJdtHpffnAI7mnSvGm4DrJf2YZJBhs5Ltaz8dp5nVpQo9i3oAcBLwhKS56bZzgeMlTSQpowtIHw2taLI9XSRLm5lVogcXEX+j847UrWU+U5lk+4hYnvUgZta7FOlRLTOzdzjZ3syKSjTODXoXODPLzz04Myuqwszoa2b2Hi5wZlZIBZvw0szs3dyDM7Oi8j04MysuFzgzKyr34MysmIJck1nWkgucmeVSqdCZnuACZ2b5ucCZWVEpGqPCucCZWT6eTcTMiqxR7sFliQ00M3sXtWdbyh5DGiPpr5KeljRP0lfT7cMk3SHpufTn0JLPnCNpvqRnJB3WXTtd4Mwsv8i4lNeRbP9+YD/g9DS9fiowKyJ2BmalrzdOtp8M/ExSU7kTuMCZWT4ZM1G7u4yNiMURMSddXwV0JNsfCUxLd5sGHJWub0i2j4gXgI5k+y65wJlZftl7cN0GP8N7ku1HdsQGpj9HpLvlTrb3IIOZ5ZLzi75lg5/hvcn2SVxql6fe2CYn25uZdUrtlRlG7SzZHljSEf4saRSwNN2eO9nel6hmlk/Wy9NuamBXyfYkCfZT0vUpwI0l24+TNEDSODYz2d5yaG+DqYfvybDt1jN12t9Z8NRArp66E2+taWLbMW9x1qXzGTi4jVWv9+XHp+zC/McGMemY1zj5ghdq3fRebfT4tzj3ihc3vN5ux/Vc98PteOy/B3HmhS+zxZbtLHm5P//39B15c3XZAbtepUIz+naVbH8hMF3SycBLwDFQ+WT7zSLpGuBTwNKI2KNa56kXt/5iFDu8by1r038EV549npO+9SK777+Su27Ylpuu2J7jzl5IvwHtfP7shbz0zEAW/n1gjVttLz/fzL8euisAffoEv57zFPfdNoRvXb2Aq7+3PU88MIhPHNfC576ylF/+cFSNW1tHqptsD3BIF5/JlWxfzUvUa0m+q1J4LYv6M2fWUA45YcmGbYueb+b9+60EYM8DV/DgrcMAaB7Yzm77rKL/gAaZb6YXmfix1Sx+sT9LX+nP6PHreOKBLQF49J7BfPSIFTVuXX2pxNdEekLVClxE3AMsr9bx68m1543lxG++SOngz5hd1/LI7ckXsB+4ZTgtiwbUqHWW1aQjX+fuPyd/Zy8+08z+hyX/QX3sUyvYdvu3a9m0+hJARLalxmo+yCDplI7vyCxrabxezew7t2bINm+z055r3rX9KxfNZ+a07fjGJz/I2tVN9O3XeL9bb9K3Xzv7fWIl99w8BIAff20Mn/7iMi6b8SxbDGqjdX2jZLn3jEo8qtUTaj7IEBFXAVcB7DWhf+1Lfk7PPLwVj9w+lEfv2pr16/qwdlUT/3nm+zjr0vl86/qnAVj0j2bmzBrazZGslj588CrmP7EFbyzrB8DC+c2ce/x4AHbYaR37HrKyls2rK57wshc54ZyXOOGclwCY999bcfOV23PWpfNZsawvQ7Zppb0d/njJaA496dUat9TKmXTUGxsuTwGGDH+bFS39kIITvrqEW64bXsPW1Zk6ufzMwgWuSu778zbMnLYdAPt8cjkHff61De+dvt9evLmqL61vi4dnDuVb1z/N6F3W1qqpvd6ALdr50MdWccnXR2/YdtBRb/DpLy4D4L7bhnD7DcNq1by61Cg9OEWVKrGk3wCTgG2AJcB3IuIX5T6z14T+cddtI8rtYnXm2NH717oJlsODMYuVsXyzbigO3np07HXgVzPte+/NX5/d3aNa1VS1HlxEHF+tY5tZbTVKD86XqGaWTwBtjVHhXODMLDf34MysuDyKamZF5R6cmRWTYwPNrKgEyIMMZlZUTrY3s2LyJaqZFVfjPIta8+mSzKzxVGrCS0nXSFoq6cmSbedJekXS3HQ5vOQ9J9ubWZVVbsLLa+l85u+LI2JiutwKTrY3s54QyShqlqXbQ+Wb+dvJ9mbWAyqcbN+JMyQ9nl7CdkzU52R7M6u+HF8T6TbZvhOXA+eTlMjzgYuAL+FkezPrEVUcRY2IDfF0kq4GbklfOtnezKosgPaMyyaQVBpAezTQMcLqZHszqy4RFXuSoXTmb0kvA98BJkmaSFJKFwCnQp0l25tZgbVXJhOwi5m/u4w2yJts7wJnZvl0XKI2ABc4M8vND9ubWXG5wJlZMTXOw/YucGaWj1O1zKzIfA/OzIrLBc7MCimAdhc4MyskDzKYWZG5wJlZIQXQ1hiPMrjAmVlOAeECZ2ZF5UtUMyskj6KaWaG5B2dmhdUgBc5TlptZPhHQ1pZt6UYXwc/DJN0h6bn059CS9xz8bGZVVt3g56nArIjYGZiVvnbws5n1kAoVuC6Cn48EpqXr04CjSrY7+NnMqimSUdQsy6YZGRGLAdKfI9LtDn42syoLiOxf9N1G0iMlr6+KiKs28cwOfjazHpD9Ua1NSbZfImlURCxOM1KXptsd/GxmVRaRxAZmWTbNTcCUdH0KcGPJdgc/m1mVVTf4+UJguqSTgZeAY5JTOvjZzHpAVDf4GeCQLvZ38LOZVZMnvDSzovLD9mZWVAFEhsew6oELnJnlE57w0swKLHyJamaF1SA9OEUdjYZIeg14sdbtqIJtgGW1boTlUtS/s3+KiG035wCSZpD8+WSxLCI2ni2kx9RVgSsqSY9swuMqVkP+OysGP6plZoXlAmdmheUC1zM2dXoYqx3/nRWA78GZWWG5B2dmheUCV0WSJqfpP/MlTa11e6x7naU8WeNygauSNO3np8Angd2B49NUIKtv1/LelCdrUC5w1bMPMD8i/hER64EbSFKBrI51kfJkDcoFrnpyJwCZWWW5wFVP7gQgM6ssF7jqyZ0AZGaV5QJXPQ8DO0saJ6k/cBxJKpCZ9RAXuCqJiFbgDGAm8DQwPSLm1bZV1p005el+YFdJL6fJTtag/CSDmRWWe3BmVlgucGZWWC5wZlZYLnBmVlgucGZWWC5wDURSm6S5kp6U9DtJAzfjWNdK+ly6/vNyEwFImiTpI5twjgWS3hNO0tX2jfZZnfNc50n6P3nbaMXmAtdY1kbExIjYA1gPnFb6ZjqDSW4R8S8R8VSZXSYBuQucWa25wDWue4H3pb2rv0q6HnhCUpOkH0p6WNLjkk4FUOIySU9J+gswouNAku6WtHe6PlnSHEmPSZolaSxJIf23tPf4MUnbSvpDeo6HJR2Qfna4pNslPSrpSjp/HvddJP1Z0mxJ8ySdstF7F6VtmSVp23TbeEkz0s/cK2m3ivxpWiE5+LkBSepLMs/cjHTTPsAeEfFCWiRWRMSHJQ0A7pN0O7AXsCvwQWAk8BRwzUbH3Ra4GjgwPdawiFgu6QpgdUT8KN3veuDiiPibpB1JntZ4P/Ad4G8R8T1JRwDvKlhd+FJ6ji2AhyX9ISJagC2BORHx75K+nR77DJKshNMi4jlJ+wI/Aw7ehD9G6wVc4BrLFpLmpuv3Ar8guXR8KCJeSLd/Atiz4/4aMATYGTgQ+E1EtAGLJN3VyfH3A+7pOFZEdDUv2seB3aUNHbStJA1Oz/HZ9LN/kfR6ht/pLElHp+tj0ra2AO3Ab9PtvwL+KGlQ+vv+ruTcAzKcw3opF7jGsjYiJpZuSP+hryndBJwZETM32u9wup+uSRn2geTWxv4RsbaTtmR+9k/SJJJiuX9EvCnpbqC5i90jPe8bG/8ZmHXF9+CKZybwFUn9ACTtImlL4B7guPQe3SjgoE4+ez/wPySNSz87LN2+Chhcst/tJJeLpPtNTFfvAb6QbvskMLSbtg4BXk+L224kPcgOfYCOXugJJJe+K4EXJB2TnkOSJnRzDuvFXOCK5+ck99fmpMEpV5L01P8EPAc8AVwO/NfGH4yI10jum/1R0mO8c4l4M3B0xyADcBawdzqI8RTvjOZ+FzhQ0hySS+WXumnrDKCvpMeB84EHSt5bA3xA0mySe2zfS7d/ATg5bd88PA28leHZRMyssNyDM7PCcoEzs8JygTOzwnKBM7PCcoEzs8JygTOzwnKBM7PCcoEzs8L6/wxMSsDLRRJjAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "run_svm = SVC(C=1.5, kernel='poly', max_iter=50, probability=True, random_state=44)\n",
    "svm_preds_pp, svm_preds, svm_cv_acc, svm_cv_prec, svm_cv_recall, svm_cv_auc = plot_and_get_metrics_scaled(best_svm)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f26d88cf",
   "metadata": {},
   "source": [
    "### Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "9858d1ef",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.64\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(class_weight='balanced', max_depth=15,\n",
       "                       min_samples_leaf=4, n_estimators=200, random_state=44)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "start_time = time.time()\n",
    "rf = RandomForestClassifier(class_weight='balanced', random_state=random_state)\n",
    "param_grid = {\n",
    "    'n_estimators': [150,200,250],\n",
    "    'criterion': ['entropy', 'gini'],\n",
    "    'min_samples_leaf': [x for x in range(2,7)],\n",
    "    'max_depth': [5,10,15,20],\n",
    "    'bootstrap': [True]\n",
    "}\n",
    "rf_clf = GridSearchCV(estimator=rf, param_grid=param_grid, scoring='precision', n_jobs=-1)\n",
    "rf_clf.fit(x_train, y_train)\n",
    "best_rf = rf_clf.best_estimator_\n",
    "print(round((time.time() - start_time)/60, 2))\n",
    "best_rf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "5c2dba97",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.53      0.53      0.53       589\n",
      "           1       0.52      0.52      0.52       570\n",
      "\n",
      "    accuracy                           0.53      1159\n",
      "   macro avg       0.53      0.53      0.53      1159\n",
      "weighted avg       0.53      0.53      0.53      1159\n",
      "\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": [
    "rf_run = RandomForestClassifier(class_weight='balanced', criterion='entropy',\n",
    "                       max_depth=5, min_samples_leaf=6, n_estimators=250,\n",
    "                       random_state=44)\n",
    "rf_preds_pp, rf_preds, rf_cv_acc, rf_cv_prec, rf_cv_recall, rf_cv_auc = plot_and_get_metrics(best_rf)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c6637ac2",
   "metadata": {},
   "source": [
    "### Light GBM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "aff428a4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "11.69\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "LGBMClassifier(class_weight='balanced', learning_rate=0.01, max_depth=10,\n",
       "               min_child_samples=2, n_estimators=150, random_state=44,\n",
       "               reg_alpha=0.1, reg_lambda=0.1)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "start_time = time.time()\n",
    "lgbmc = lgb.LGBMClassifier(class_weight='balanced', random_state=random_state)\n",
    "\n",
    "param_grid = {\n",
    "    'max_depth': [5,10,15,20],\n",
    "    # 'num_leaves': [5, 10],\n",
    "    'learning_rate': [0.001, 0.01, 0.025],\n",
    "    'n_estimators': [150, 200, 250],\n",
    "    # 'feature_fraction': [0.5],\n",
    "    'min_child_samples': [x for x in range(2,5)],\n",
    "    'reg_alpha': [0.1, 0.5,1,1.5],\n",
    "    'reg_lambda': [0.1, 0.5,1,1.5]\n",
    "}\n",
    "\n",
    "lgbmc_clf = GridSearchCV(estimator=lgbmc,\n",
    "                         param_grid=param_grid,\n",
    "                         scoring='accuracy',\n",
    "                         cv=5,\n",
    "                         verbose=False,\n",
    "                         n_jobs=-1)\n",
    "\n",
    "lgbmc_clf.fit(x_train, y_train)\n",
    "best_lgb = lgbmc_clf.best_estimator_\n",
    "print(round((time.time() - start_time)/60, 2))\n",
    "best_lgb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "4e4ff2ff",
   "metadata": {},
   "outputs": [],
   "source": [
    "lgbm_run = lgb.LGBMClassifier(class_weight='balanced', learning_rate=0.025, max_depth=10,\n",
    "               min_child_samples=2, n_estimators=200, random_state=44,\n",
    "               reg_alpha=0.1, reg_lambda=0.5)\n",
    "best_lgb = lgbm_run.fit(x_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "6f1b511d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.52      0.51      0.52       589\n",
      "           1       0.50      0.51      0.51       570\n",
      "\n",
      "    accuracy                           0.51      1159\n",
      "   macro avg       0.51      0.51      0.51      1159\n",
      "weighted avg       0.51      0.51      0.51      1159\n",
      "\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": [
    "lgb_preds_pp, lgb_preds, lgb_cv_acc, lgb_cv_prec, lgb_cv_recall, lgb_cv_auc = plot_and_get_metrics(best_lgb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "5c3f99dd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1440 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20,20))\n",
    "sns.barplot(data=pd.DataFrame({'Feature': x_train.columns, 'Importance': best_lgb.feature_importances_}).\\\n",
    "    sort_values('Importance', ascending=False), x='Importance', y='Feature', color='g').set_title('LGBM')\n",
    "plt.savefig('feat_import_LGBM.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7b1378d6",
   "metadata": {},
   "source": [
    "### XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "634a4999",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[23:05:53] WARNING: D:\\bld\\xgboost-split_1645118015404\\work\\src\\learner.cc:1115: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'binary:logistic' was changed from 'error' to 'logloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n",
      "9.392348690827687\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "XGBClassifier(alpha=0.1, base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "              colsample_bynode=1, colsample_bytree=1, enable_categorical=False,\n",
       "              gamma=0, gpu_id=-1, importance_type=None,\n",
       "              interaction_constraints='', lambda=1, learning_rate=0.05,\n",
       "              max_delta_step=0, max_depth=20, min_child_weight=1, missing=nan,\n",
       "              monotone_constraints='()', n_estimators=300, n_jobs=8,\n",
       "              num_parallel_tree=1, predictor='auto', random_state=44,\n",
       "              reg_alpha=0.100000001, reg_lambda=1, scale_pos_weight=4,\n",
       "              subsample=0.75, tree_method='exact', validate_parameters=1,\n",
       "              verbosity=None)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "xgb = XGBClassifier(scale_pos_weight=4, random_state=random_state)\n",
    "start_time = time.time()\n",
    "param_grid = {\n",
    "    'n_estimators': [150,250,300],\n",
    "    'learning_rate': [0.05,0.1],\n",
    "    'max_depth': [20,25],\n",
    "    'subsample': [0.5, 0.75],\n",
    "    'alpha': [0.1,1,1.5],\n",
    "    'lambda': [0.5,1,5]\n",
    "}\n",
    "\n",
    "\n",
    "xgb_clf = GridSearchCV(estimator=xgb, param_grid=param_grid, scoring='accuracy', n_jobs=-1)\n",
    "\n",
    "xgb_clf.fit(x_train, y_train)\n",
    "best_xgb = xgb_clf.best_estimator_\n",
    "print((time.time() - start_time)/60)\n",
    "best_xgb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "1216ff7a",
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb_run = XGBClassifier(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
    "              colsample_bynode=1, colsample_bytree=0.8,\n",
    "              enable_categorical=False, eval_metric='error', gamma=0.5,\n",
    "              gpu_id=-1, importance_type=None, interaction_constraints='',\n",
    "              learning_rate=0.025, max_delta_step=0, max_depth=25,\n",
    "              min_child_weight=1, monotone_constraints='()',\n",
    "              n_estimators=200, n_jobs=8, num_parallel_tree=1, predictor='auto',\n",
    "              random_state=44, reg_alpha=0, reg_lambda=1, scale_pos_weight=4,\n",
    "              subsample=0.75, tree_method='exact', validate_parameters=1,\n",
    "              verbosity=None)\n",
    "best_xgb = xgb_run.fit(x_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "0465d64e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.52      0.25      0.34       589\n",
      "           1       0.49      0.76      0.60       570\n",
      "\n",
      "    accuracy                           0.50      1159\n",
      "   macro avg       0.51      0.50      0.47      1159\n",
      "weighted avg       0.51      0.50      0.47      1159\n",
      "\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": [
    "xgb_preds_pp, xgb_preds, xgb_cv_acc, xgb_cv_prec, xgb_cv_recall, xgb_cv_auc = plot_and_get_metrics(best_xgb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "cabf6abf",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1440 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20,20))\n",
    "sns.barplot(data=pd.DataFrame({'Feature': x_train.columns, 'Importance': best_xgb.feature_importances_}).\\\n",
    "    sort_values('Importance', ascending=False), x='Importance', y='Feature', color='g').set_title('XGB')\n",
    "plt.savefig('feat_import_xgb.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "047afcb6",
   "metadata": {},
   "outputs": [],
   "source": [
    "model_names = ['XGB', 'LGBM', 'RF',  'KNN', 'SVM','NB']\n",
    "cv_accuracy = pd.Series([xgb_cv_acc, lgb_cv_acc, rf_cv_acc, knn_cv_acc, svm_cv_acc, nb_cv_acc], name='Accuracy')\n",
    "cv_recall = pd.Series([xgb_cv_recall, lgb_cv_recall, rf_cv_recall, knn_cv_recall, svm_cv_recall, nb_cv_recall], name='Recall')\n",
    "cv_precision = pd.Series([xgb_cv_prec, lgb_cv_prec, rf_cv_prec, knn_cv_prec, svm_cv_prec, nb_cv_prec],name='Precision')\n",
    "cv_auc = pd.Series([xgb_cv_auc, lgb_cv_auc, rf_cv_auc,  knn_cv_auc, svm_cv_auc, nb_cv_auc], name='AUC')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "06636609",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_c3f69_row0_col1, #T_c3f69_row2_col0, #T_c3f69_row2_col2, #T_c3f69_row2_col3 {\n",
       "  background-color: green;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_c3f69_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th class=\"col_heading level0 col0\" >Accuracy</th>\n",
       "      <th class=\"col_heading level0 col1\" >Recall</th>\n",
       "      <th class=\"col_heading level0 col2\" >Precision</th>\n",
       "      <th class=\"col_heading level0 col3\" >AUC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_c3f69_level0_row0\" class=\"row_heading level0 row0\" >XGB</th>\n",
       "      <td id=\"T_c3f69_row0_col0\" class=\"data row0 col0\" >0.500431</td>\n",
       "      <td id=\"T_c3f69_row0_col1\" class=\"data row0 col1\" >0.761404</td>\n",
       "      <td id=\"T_c3f69_row0_col2\" class=\"data row0 col2\" >0.494869</td>\n",
       "      <td id=\"T_c3f69_row0_col3\" class=\"data row0 col3\" >0.504641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c3f69_level0_row1\" class=\"row_heading level0 row1\" >LGBM</th>\n",
       "      <td id=\"T_c3f69_row1_col0\" class=\"data row1 col0\" >0.511648</td>\n",
       "      <td id=\"T_c3f69_row1_col1\" class=\"data row1 col1\" >0.512281</td>\n",
       "      <td id=\"T_c3f69_row1_col2\" class=\"data row1 col2\" >0.503448</td>\n",
       "      <td id=\"T_c3f69_row1_col3\" class=\"data row1 col3\" >0.511658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c3f69_level0_row2\" class=\"row_heading level0 row2\" >RF</th>\n",
       "      <td id=\"T_c3f69_row2_col0\" class=\"data row2 col0\" >0.525453</td>\n",
       "      <td id=\"T_c3f69_row2_col1\" class=\"data row2 col1\" >0.524561</td>\n",
       "      <td id=\"T_c3f69_row2_col2\" class=\"data row2 col2\" >0.517301</td>\n",
       "      <td id=\"T_c3f69_row2_col3\" class=\"data row2 col3\" >0.525439</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c3f69_level0_row3\" class=\"row_heading level0 row3\" >KNN</th>\n",
       "      <td id=\"T_c3f69_row3_col0\" class=\"data row3 col0\" >0.505608</td>\n",
       "      <td id=\"T_c3f69_row3_col1\" class=\"data row3 col1\" >0.459649</td>\n",
       "      <td id=\"T_c3f69_row3_col2\" class=\"data row3 col2\" >0.497154</td>\n",
       "      <td id=\"T_c3f69_row3_col3\" class=\"data row3 col3\" >0.504867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c3f69_level0_row4\" class=\"row_heading level0 row4\" >SVM</th>\n",
       "      <td id=\"T_c3f69_row4_col0\" class=\"data row4 col0\" >0.500431</td>\n",
       "      <td id=\"T_c3f69_row4_col1\" class=\"data row4 col1\" >0.138596</td>\n",
       "      <td id=\"T_c3f69_row4_col2\" class=\"data row4 col2\" >0.473054</td>\n",
       "      <td id=\"T_c3f69_row4_col3\" class=\"data row4 col3\" >0.494595</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c3f69_level0_row5\" class=\"row_heading level0 row5\" >NB</th>\n",
       "      <td id=\"T_c3f69_row5_col0\" class=\"data row5 col0\" >0.493529</td>\n",
       "      <td id=\"T_c3f69_row5_col1\" class=\"data row5 col1\" >0.664912</td>\n",
       "      <td id=\"T_c3f69_row5_col2\" class=\"data row5 col2\" >0.489032</td>\n",
       "      <td id=\"T_c3f69_row5_col3\" class=\"data row5 col3\" >0.496293</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1f418f2b910>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scores_summary = pd.concat([cv_accuracy, cv_recall, cv_precision, cv_auc], axis=1)\n",
    "scores_summary.columns = ['Accuracy', 'Recall', 'Precision', 'AUC']\n",
    "scores_summary.index = model_names\n",
    "scores_summary.style.highlight_max(color='green', axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "798b1e8b",
   "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>true</th>\n",
       "      <th>knn_preds_pp</th>\n",
       "      <th>rf_preds_pp</th>\n",
       "      <th>lgb_preds_pp</th>\n",
       "      <th>xgb_preds_pp</th>\n",
       "      <th>nb_preds_pp</th>\n",
       "      <th>svm_preds_pp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1524</th>\n",
       "      <td>0</td>\n",
       "      <td>0.574237</td>\n",
       "      <td>0.559987</td>\n",
       "      <td>0.682274</td>\n",
       "      <td>0.771202</td>\n",
       "      <td>0.635564</td>\n",
       "      <td>0.493167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>218</th>\n",
       "      <td>1</td>\n",
       "      <td>0.459179</td>\n",
       "      <td>0.361550</td>\n",
       "      <td>0.302648</td>\n",
       "      <td>0.255885</td>\n",
       "      <td>0.006237</td>\n",
       "      <td>0.457371</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1039</th>\n",
       "      <td>0</td>\n",
       "      <td>0.617017</td>\n",
       "      <td>0.453392</td>\n",
       "      <td>0.288789</td>\n",
       "      <td>0.666402</td>\n",
       "      <td>0.791332</td>\n",
       "      <td>0.517977</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1229</th>\n",
       "      <td>0</td>\n",
       "      <td>0.652599</td>\n",
       "      <td>0.443908</td>\n",
       "      <td>0.203940</td>\n",
       "      <td>0.301594</td>\n",
       "      <td>0.738815</td>\n",
       "      <td>0.552466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>1</td>\n",
       "      <td>0.586783</td>\n",
       "      <td>0.512780</td>\n",
       "      <td>0.739077</td>\n",
       "      <td>0.500567</td>\n",
       "      <td>0.723960</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      true  knn_preds_pp  rf_preds_pp  lgb_preds_pp  xgb_preds_pp  \\\n",
       "1524     0      0.574237     0.559987      0.682274      0.771202   \n",
       "218      1      0.459179     0.361550      0.302648      0.255885   \n",
       "1039     0      0.617017     0.453392      0.288789      0.666402   \n",
       "1229     0      0.652599     0.443908      0.203940      0.301594   \n",
       "61       1      0.586783     0.512780      0.739077      0.500567   \n",
       "\n",
       "      nb_preds_pp  svm_preds_pp  \n",
       "1524     0.635564      0.493167  \n",
       "218      0.006237      0.457371  \n",
       "1039     0.791332      0.517977  \n",
       "1229     0.738815      0.552466  \n",
       "61       0.723960      0.500000  "
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "comp_df = pd.DataFrame({'true': y_train, 'knn_preds_pp': knn_preds_pp, 'rf_preds_pp': rf_preds_pp,\\\n",
    "                        'lgb_preds_pp':lgb_preds_pp,'xgb_preds_pp':xgb_preds_pp ,'nb_preds_pp':nb_preds_pp,\\\n",
    "                        'svm_preds_pp': svm_preds_pp})\n",
    "comp_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b21b6fef",
   "metadata": {},
   "source": [
    "Only bet when our model probability is higher than Vegas' implied probability. If both higher than Vegas, take the one with higher margin."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "065438d9",
   "metadata": {},
   "outputs": [],
   "source": [
    "def under_return(df):\n",
    "    if (df['bet_under'] and (df['true'] == 1)):\n",
    "        return -100\n",
    "    elif (df['bet_under'] & (df['true'] == 0)):\n",
    "        #standard offer odds of -110 on the bet\n",
    "        return round((100/-110)*-100,2)\n",
    "    else: \n",
    "        return 0\n",
    "    \n",
    "def over_return(df):\n",
    "    if (df['bet_over'] and (df['true'] == 0)):\n",
    "        return -100\n",
    "    elif (df['bet_over'] & (df['true'] == 1)):\n",
    "        return round((100/-110)*-100,2)\n",
    "    else: \n",
    "        return 0\n",
    "def pred_vs_vegas(model_preds, comp, margin = 0):\n",
    "    '''\n",
    "    \n",
    "    :param margin: margin between our predictions and vegas predictions to trigger making a bet (example: 0.05)\n",
    "    :param comp: comparison dataframe to use for function (train or test predictions)\n",
    "    :param model_preds: model column name of prediction probability\n",
    "    :return:% return, $ risked, total $ returned\n",
    "    '''\n",
    "    #create working dataframe\n",
    "    df_w = comp[['true',model_preds]]\n",
    "    \n",
    "    #calculate differential of our predictions vs standard -110 offer (translates to 52.38% probability)\n",
    "    df_w['over_prob_adv'] = df_w[model_preds] - 0.5238\n",
    "    df_w['under_prob_adv'] = (1-df_w[model_preds]) - 0.5238\n",
    "    \n",
    "    #flag whether the home differential is higher than the away differential\n",
    "    df_w['over_adv_higher'] = df_w['over_prob_adv'] > df_w['under_prob_adv']\n",
    "    \n",
    "    #flag whether to bet on the home or away team\n",
    "    df_w['bet_over'] = np.where((df_w['over_adv_higher'] & (df_w['over_prob_adv']>margin)), True, False)\n",
    "    df_w['bet_under'] = np.where(((df_w['bet_over'] == False) & (df_w['under_prob_adv']>margin)), True, False)\n",
    "    \n",
    "    #generate bet returns\n",
    "    df_w['over_risk'] = np.where(df_w['bet_over'], 100, 0)\n",
    "    df_w['under_risk'] = np.where(df_w['bet_under'], 100, 0)\n",
    "    df_w['under_return'] = df_w.apply(under_return, axis = 1)\n",
    "    df_w['over_return'] = df_w.apply(over_return, axis = 1)\n",
    "    \n",
    "    #aggregate risks and returns\n",
    "    df_w['risk'] = df_w['over_risk'] + df_w['under_risk']\n",
    "    df_w['return'] = df_w['over_return'] + df_w['under_return']\n",
    "    \n",
    "    return df_w\n",
    "\n",
    "def aggregate_returns(df_w):\n",
    "    return df_w['risk'].sum(),df_w['return'].sum(),(df_w['return'].sum()/df_w['risk'].sum())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "e651cc63",
   "metadata": {},
   "outputs": [],
   "source": [
    "final_models = ['knn_preds_pp','svm_preds_pp','nb_preds_pp','rf_preds_pp','xgb_preds_pp','lgb_preds_pp']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "ee5ff2f3",
   "metadata": {},
   "outputs": [],
   "source": [
    "mod_risk = []\n",
    "mod_ret = []\n",
    "mod_pct_ret = []\n",
    "models = []\n",
    "margins = []\n",
    "for model in final_models:\n",
    "    for margin in np.arange(0,.2,0.01):\n",
    "        df = pred_vs_vegas(model,comp_df,margin)\n",
    "        risk, ret, pct = aggregate_returns(df)\n",
    "        mod_risk.append(risk)\n",
    "        mod_ret.append(ret)\n",
    "        mod_pct_ret.append(pct)\n",
    "        models.append(model)\n",
    "        margins.append(margin)\n",
    "\n",
    "higher_exp_val_df = pd.DataFrame({'model':models,'margin': margins, 'pct_ret': mod_pct_ret, 'risked':mod_risk, 'dollar_ret':mod_ret})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "160c09de",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,6))\n",
    "sns.lineplot(x='margin', y='pct_ret', hue='model', data=higher_exp_val_df)\n",
    "plt.suptitle('% Return on O/U Bets per Model by Margin', size=16)\n",
    "plt.legend(loc='lower left')\n",
    "plt.savefig('strat_3_ret.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "b679dc93",
   "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>model</th>\n",
       "      <th>margin</th>\n",
       "      <th>pct_ret</th>\n",
       "      <th>risked</th>\n",
       "      <th>dollar_ret</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.042796</td>\n",
       "      <td>107900</td>\n",
       "      <td>-4617.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.01</td>\n",
       "      <td>-0.036376</td>\n",
       "      <td>105200</td>\n",
       "      <td>-3826.79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.02</td>\n",
       "      <td>-0.035186</td>\n",
       "      <td>102300</td>\n",
       "      <td>-3599.53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.03</td>\n",
       "      <td>-0.028942</td>\n",
       "      <td>98300</td>\n",
       "      <td>-2845.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>84</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.04</td>\n",
       "      <td>-0.027364</td>\n",
       "      <td>95000</td>\n",
       "      <td>-2599.56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.05</td>\n",
       "      <td>-0.033080</td>\n",
       "      <td>92600</td>\n",
       "      <td>-3063.21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.06</td>\n",
       "      <td>-0.042258</td>\n",
       "      <td>89700</td>\n",
       "      <td>-3790.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.07</td>\n",
       "      <td>-0.059796</td>\n",
       "      <td>86500</td>\n",
       "      <td>-5172.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>88</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.08</td>\n",
       "      <td>-0.056868</td>\n",
       "      <td>83600</td>\n",
       "      <td>-4754.17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.09</td>\n",
       "      <td>-0.060942</td>\n",
       "      <td>80100</td>\n",
       "      <td>-4881.46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.10</td>\n",
       "      <td>-0.066774</td>\n",
       "      <td>76100</td>\n",
       "      <td>-5081.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.11</td>\n",
       "      <td>-0.066662</td>\n",
       "      <td>72000</td>\n",
       "      <td>-4799.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.12</td>\n",
       "      <td>-0.060779</td>\n",
       "      <td>68500</td>\n",
       "      <td>-4163.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.13</td>\n",
       "      <td>-0.070454</td>\n",
       "      <td>64900</td>\n",
       "      <td>-4572.44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.14</td>\n",
       "      <td>-0.069847</td>\n",
       "      <td>62600</td>\n",
       "      <td>-4372.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.15</td>\n",
       "      <td>-0.075829</td>\n",
       "      <td>59700</td>\n",
       "      <td>-4527.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.16</td>\n",
       "      <td>-0.068813</td>\n",
       "      <td>57200</td>\n",
       "      <td>-3936.11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.17</td>\n",
       "      <td>-0.086641</td>\n",
       "      <td>53300</td>\n",
       "      <td>-4617.95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.18</td>\n",
       "      <td>-0.090725</td>\n",
       "      <td>50600</td>\n",
       "      <td>-4590.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>xgb_preds_pp</td>\n",
       "      <td>0.19</td>\n",
       "      <td>-0.076876</td>\n",
       "      <td>48600</td>\n",
       "      <td>-3736.15</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           model  margin   pct_ret  risked  dollar_ret\n",
       "80  xgb_preds_pp    0.00 -0.042796  107900    -4617.69\n",
       "81  xgb_preds_pp    0.01 -0.036376  105200    -3826.79\n",
       "82  xgb_preds_pp    0.02 -0.035186  102300    -3599.53\n",
       "83  xgb_preds_pp    0.03 -0.028942   98300    -2845.00\n",
       "84  xgb_preds_pp    0.04 -0.027364   95000    -2599.56\n",
       "85  xgb_preds_pp    0.05 -0.033080   92600    -3063.21\n",
       "86  xgb_preds_pp    0.06 -0.042258   89700    -3790.50\n",
       "87  xgb_preds_pp    0.07 -0.059796   86500    -5172.34\n",
       "88  xgb_preds_pp    0.08 -0.056868   83600    -4754.17\n",
       "89  xgb_preds_pp    0.09 -0.060942   80100    -4881.46\n",
       "90  xgb_preds_pp    0.10 -0.066774   76100    -5081.48\n",
       "91  xgb_preds_pp    0.11 -0.066662   72000    -4799.68\n",
       "92  xgb_preds_pp    0.12 -0.060779   68500    -4163.33\n",
       "93  xgb_preds_pp    0.13 -0.070454   64900    -4572.44\n",
       "94  xgb_preds_pp    0.14 -0.069847   62600    -4372.45\n",
       "95  xgb_preds_pp    0.15 -0.075829   59700    -4527.01\n",
       "96  xgb_preds_pp    0.16 -0.068813   57200    -3936.11\n",
       "97  xgb_preds_pp    0.17 -0.086641   53300    -4617.95\n",
       "98  xgb_preds_pp    0.18 -0.090725   50600    -4590.69\n",
       "99  xgb_preds_pp    0.19 -0.076876   48600    -3736.15"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "higher_exp_val_df.loc[higher_exp_val_df['model'] == 'xgb_preds_pp']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "4c675d6c",
   "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>true</th>\n",
       "      <th>knn_preds_pp</th>\n",
       "      <th>rf_preds_pp</th>\n",
       "      <th>lgb_preds_pp</th>\n",
       "      <th>xgb_preds_pp</th>\n",
       "      <th>nb_preds_pp</th>\n",
       "      <th>svm_preds_pp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1552</th>\n",
       "      <td>0</td>\n",
       "      <td>0.561051</td>\n",
       "      <td>0.654624</td>\n",
       "      <td>0.582932</td>\n",
       "      <td>0.858019</td>\n",
       "      <td>0.592908</td>\n",
       "      <td>0.490725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1553</th>\n",
       "      <td>0</td>\n",
       "      <td>0.330413</td>\n",
       "      <td>0.350331</td>\n",
       "      <td>0.268958</td>\n",
       "      <td>0.281931</td>\n",
       "      <td>0.499178</td>\n",
       "      <td>0.494435</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1554</th>\n",
       "      <td>0</td>\n",
       "      <td>0.467447</td>\n",
       "      <td>0.426394</td>\n",
       "      <td>0.213005</td>\n",
       "      <td>0.414488</td>\n",
       "      <td>0.701843</td>\n",
       "      <td>0.493939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1555</th>\n",
       "      <td>1</td>\n",
       "      <td>0.442951</td>\n",
       "      <td>0.445769</td>\n",
       "      <td>0.492713</td>\n",
       "      <td>0.618574</td>\n",
       "      <td>0.665853</td>\n",
       "      <td>0.494377</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1556</th>\n",
       "      <td>0</td>\n",
       "      <td>0.532872</td>\n",
       "      <td>0.514398</td>\n",
       "      <td>0.575656</td>\n",
       "      <td>0.823464</td>\n",
       "      <td>0.689450</td>\n",
       "      <td>0.493043</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      true  knn_preds_pp  rf_preds_pp  lgb_preds_pp  xgb_preds_pp  \\\n",
       "1552     0      0.561051     0.654624      0.582932      0.858019   \n",
       "1553     0      0.330413     0.350331      0.268958      0.281931   \n",
       "1554     0      0.467447     0.426394      0.213005      0.414488   \n",
       "1555     1      0.442951     0.445769      0.492713      0.618574   \n",
       "1556     0      0.532872     0.514398      0.575656      0.823464   \n",
       "\n",
       "      nb_preds_pp  svm_preds_pp  \n",
       "1552     0.592908      0.490725  \n",
       "1553     0.499178      0.494435  \n",
       "1554     0.701843      0.493939  \n",
       "1555     0.665853      0.494377  \n",
       "1556     0.689450      0.493043  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "comp_df2 = pd.DataFrame({'true': games_2022['home_win'], 'knn_preds_pp': best_knn.predict_proba(games_2022.filter(pred_cols).fillna(0))[:,1],\\\n",
    "                        'rf_preds_pp': best_rf.predict_proba(games_2022.filter(pred_cols).fillna(0))[:,1],\\\n",
    "                        'lgb_preds_pp':best_lgb.predict_proba(games_2022.filter(pred_cols).fillna(0))[:,1],\\\n",
    "                        'xgb_preds_pp':best_xgb.predict_proba(games_2022.filter(pred_cols).fillna(0))[:,1],\\\n",
    "                        'nb_preds_pp':best_nb.predict_proba(games_2022.filter(pred_cols).fillna(0))[:,1],\\\n",
    "                        'svm_preds_pp':best_svm.predict_proba(scaler.transform(games_2022.filter(pred_cols).fillna(0)))[:,1]})\n",
    "comp_df2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "60107aab",
   "metadata": {},
   "outputs": [],
   "source": [
    "mod_risk = []\n",
    "mod_ret = []\n",
    "mod_pct_ret = []\n",
    "models = []\n",
    "margins = []\n",
    "for model in final_models:\n",
    "    for margin in np.arange(0,.2,0.01):\n",
    "        df = pred_vs_vegas(model,comp_df2,margin)\n",
    "        risk, ret, pct = aggregate_returns(df)\n",
    "        mod_risk.append(risk)\n",
    "        mod_ret.append(ret)\n",
    "        mod_pct_ret.append(pct)\n",
    "        models.append(model)\n",
    "        margins.append(margin)\n",
    "\n",
    "higher_exp_val_df2 = pd.DataFrame({'model':models,'margin': margins, 'pct_ret': mod_pct_ret, 'risked':mod_risk, 'dollar_ret':mod_ret})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "fc3aeaf8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,6))\n",
    "sns.lineplot(x='margin', y='pct_ret', hue='model', data=higher_exp_val_df2)\n",
    "plt.suptitle('% Return on O/U per Model by Margin in 2022', size=16)\n",
    "plt.legend(loc='upper left')\n",
    "plt.savefig('strat_3_ret.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "8bb132a0",
   "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>model</th>\n",
       "      <th>margin</th>\n",
       "      <th>pct_ret</th>\n",
       "      <th>risked</th>\n",
       "      <th>dollar_ret</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.01</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.02</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.03</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.04</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.05</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.06</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.07</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.08</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.10</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.11</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.12</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.13</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.14</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.15</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.16</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.18</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>svm_preds_pp</td>\n",
       "      <td>0.19</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           model  margin  pct_ret  risked  dollar_ret\n",
       "20  svm_preds_pp    0.00      NaN       0         0.0\n",
       "21  svm_preds_pp    0.01      NaN       0         0.0\n",
       "22  svm_preds_pp    0.02      NaN       0         0.0\n",
       "23  svm_preds_pp    0.03      NaN       0         0.0\n",
       "24  svm_preds_pp    0.04      NaN       0         0.0\n",
       "25  svm_preds_pp    0.05      NaN       0         0.0\n",
       "26  svm_preds_pp    0.06      NaN       0         0.0\n",
       "27  svm_preds_pp    0.07      NaN       0         0.0\n",
       "28  svm_preds_pp    0.08      NaN       0         0.0\n",
       "29  svm_preds_pp    0.09      NaN       0         0.0\n",
       "30  svm_preds_pp    0.10      NaN       0         0.0\n",
       "31  svm_preds_pp    0.11      NaN       0         0.0\n",
       "32  svm_preds_pp    0.12      NaN       0         0.0\n",
       "33  svm_preds_pp    0.13      NaN       0         0.0\n",
       "34  svm_preds_pp    0.14      NaN       0         0.0\n",
       "35  svm_preds_pp    0.15      NaN       0         0.0\n",
       "36  svm_preds_pp    0.16      NaN       0         0.0\n",
       "37  svm_preds_pp    0.17      NaN       0         0.0\n",
       "38  svm_preds_pp    0.18      NaN       0         0.0\n",
       "39  svm_preds_pp    0.19      NaN       0         0.0"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "higher_exp_val_df2.loc[higher_exp_val_df2['model'] == 'svm_preds_pp']"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}