Daniels Delirium.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "32cb1a55",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np, pandas as pd\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.impute import KNNImputer, SimpleImputer\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.model_selection import cross_validate\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "from tensorflow.keras.layers import *\n",
    "from tensorflow.keras.models import *\n",
    "from tensorflow.keras.wrappers.scikit_learn import KerasClassifier\n",
    "from tensorflow import keras\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "46b5253e",
   "metadata": {},
   "outputs": [],
   "source": [
    "trainData = pd.read_csv(\"Data/train.csv\")\n",
    "testData = pd.read_csv(\"Data/test.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "2248a414",
   "metadata": {},
   "outputs": [],
   "source": [
    "trainDataSmall = trainData[trainData.columns[9900:10002]]\n",
    "testDataSmall = testData[testData.columns[9900:10001]]\n",
    "\n",
    "train_df = trainData\n",
    "test_df = testData\n",
    "\n",
    "trainDataClass = train_df[\"Outcome\"]\n",
    "trainDataIds = train_df[\"Id\"]\n",
    "trainDataFeatures = train_df.drop(columns = [\"Outcome\", \"Id\"])\n",
    "\n",
    "testDataIds = test_df[\"Id\"]\n",
    "testDataFeatures = test_df.drop(columns = [\"Id\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "da8a33b7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 371397 NaNs\n",
      "Class outcome 0 accounts for 16.7%\n",
      "Class outcome 1 accounts for 20.2%\n",
      "Class outcome 2 accounts for 25.4%\n",
      "Class outcome 3 accounts for 37.7%\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>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Gene 1</th>\n",
       "      <td>1427.0</td>\n",
       "      <td>203.563420</td>\n",
       "      <td>59.951166</td>\n",
       "      <td>0.0</td>\n",
       "      <td>162.00</td>\n",
       "      <td>203.0</td>\n",
       "      <td>245.00</td>\n",
       "      <td>394.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Gene 2</th>\n",
       "      <td>1424.0</td>\n",
       "      <td>43497.110253</td>\n",
       "      <td>13257.898487</td>\n",
       "      <td>0.0</td>\n",
       "      <td>34875.75</td>\n",
       "      <td>43697.0</td>\n",
       "      <td>52267.25</td>\n",
       "      <td>90978.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Gene 3</th>\n",
       "      <td>1424.0</td>\n",
       "      <td>330.366573</td>\n",
       "      <td>90.843454</td>\n",
       "      <td>0.0</td>\n",
       "      <td>270.00</td>\n",
       "      <td>331.0</td>\n",
       "      <td>391.00</td>\n",
       "      <td>606.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Gene 4</th>\n",
       "      <td>1428.0</td>\n",
       "      <td>241.135154</td>\n",
       "      <td>65.468817</td>\n",
       "      <td>0.0</td>\n",
       "      <td>198.75</td>\n",
       "      <td>239.0</td>\n",
       "      <td>286.00</td>\n",
       "      <td>493.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Gene 5</th>\n",
       "      <td>1436.0</td>\n",
       "      <td>1217.438022</td>\n",
       "      <td>348.133107</td>\n",
       "      <td>90.0</td>\n",
       "      <td>975.00</td>\n",
       "      <td>1207.5</td>\n",
       "      <td>1446.25</td>\n",
       "      <td>2302.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Gene 3998</th>\n",
       "      <td>1415.0</td>\n",
       "      <td>31343.093993</td>\n",
       "      <td>9077.064820</td>\n",
       "      <td>0.0</td>\n",
       "      <td>25184.00</td>\n",
       "      <td>31522.0</td>\n",
       "      <td>37223.50</td>\n",
       "      <td>64221.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Gene 3999</th>\n",
       "      <td>1422.0</td>\n",
       "      <td>5.962729</td>\n",
       "      <td>1.993484</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.00</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.00</td>\n",
       "      <td>12.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Gene 4000</th>\n",
       "      <td>1432.0</td>\n",
       "      <td>54788.787709</td>\n",
       "      <td>14358.277174</td>\n",
       "      <td>9460.0</td>\n",
       "      <td>44698.75</td>\n",
       "      <td>55209.0</td>\n",
       "      <td>64311.25</td>\n",
       "      <td>97776.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Outcome</th>\n",
       "      <td>1520.0</td>\n",
       "      <td>1.840789</td>\n",
       "      <td>1.105814</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.00</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Id</th>\n",
       "      <td>1520.0</td>\n",
       "      <td>981.832237</td>\n",
       "      <td>576.310115</td>\n",
       "      <td>1.0</td>\n",
       "      <td>482.75</td>\n",
       "      <td>958.5</td>\n",
       "      <td>1481.25</td>\n",
       "      <td>1996.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4002 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            count          mean           std     min       25%      50%  \\\n",
       "Gene 1     1427.0    203.563420     59.951166     0.0    162.00    203.0   \n",
       "Gene 2     1424.0  43497.110253  13257.898487     0.0  34875.75  43697.0   \n",
       "Gene 3     1424.0    330.366573     90.843454     0.0    270.00    331.0   \n",
       "Gene 4     1428.0    241.135154     65.468817     0.0    198.75    239.0   \n",
       "Gene 5     1436.0   1217.438022    348.133107    90.0    975.00   1207.5   \n",
       "...           ...           ...           ...     ...       ...      ...   \n",
       "Gene 3998  1415.0  31343.093993   9077.064820     0.0  25184.00  31522.0   \n",
       "Gene 3999  1422.0      5.962729      1.993484     0.0      5.00      6.0   \n",
       "Gene 4000  1432.0  54788.787709  14358.277174  9460.0  44698.75  55209.0   \n",
       "Outcome    1520.0      1.840789      1.105814     0.0      1.00      2.0   \n",
       "Id         1520.0    981.832237    576.310115     1.0    482.75    958.5   \n",
       "\n",
       "                75%      max  \n",
       "Gene 1       245.00    394.0  \n",
       "Gene 2     52267.25  90978.0  \n",
       "Gene 3       391.00    606.0  \n",
       "Gene 4       286.00    493.0  \n",
       "Gene 5      1446.25   2302.0  \n",
       "...             ...      ...  \n",
       "Gene 3998  37223.50  64221.0  \n",
       "Gene 3999      7.00     12.0  \n",
       "Gene 4000  64311.25  97776.0  \n",
       "Outcome        3.00      3.0  \n",
       "Id          1481.25   1996.0  \n",
       "\n",
       "[4002 rows x 8 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# exercise 2\n",
    "\n",
    "print(f\"There are {trainData.isna().sum().sum()} NaNs\")\n",
    "outcomes = trainData[\"Outcome\"]\n",
    "for i in range(4):\n",
    "   print(f\"Class outcome {i} accounts for {(outcomes == i).sum()/len(trainData)*100:.1f}%\")\n",
    "summary = trainData.describe()\n",
    "display(summary.T)\n",
    "sns.scatterplot(data = summary.T, x = \"count\", y = \"max\")\n",
    "plt.show()\n",
    "sns.scatterplot(data = summary.T, x = \"count\", y = \"std\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "id": "7d7d9925",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "trainData0 = trainData[trainData[\"Outcome\"] == 0]\n",
    "trainData1 = trainData[trainData[\"Outcome\"] == 1]\n",
    "trainData2 = trainData[trainData[\"Outcome\"] == 2]\n",
    "trainData3 = trainData[trainData[\"Outcome\"] == 3]\n",
    "\n",
    "datasets = [trainData0, trainData1, trainData2, trainData3]\n",
    "\n",
    "leastAmount = (trainData[trainData[\"Outcome\"] == 0].shape)[0]\n",
    "\n",
    "changedSets = [trainData0, None, None, None]\n",
    "for n in range(1,4):\n",
    "    dataset = datasets[n]\n",
    "    dataset.index = range(dataset.shape[0])\n",
    "    indices = np.random.choice(dataset.shape[0], leastAmount, replace=False)\n",
    "    dataset = dataset.iloc[indices]\n",
    "    dataset.index = range(dataset.shape[0])\n",
    "    changedSets[n] = dataset\n",
    "datasets = changedSets\n",
    "\n",
    "trainDataDownSampled = pd.concat([datasets[0], datasets[1], datasets[2], datasets[3]], axis=0)\n",
    "\n",
    "\n",
    "# print(f\"There are {trainDataDownSampled.isna().sum().sum()} NaNs\")\n",
    "# outcomes = trainDataDownSampled[\"Outcome\"]\n",
    "# for i in range(4):\n",
    "#    print(f\"Class outcome {i} accounts for {(outcomes == i).sum()/len(trainDataDownSampled)*100:.1f}%\")\n",
    "# summary = trainDataDownSampled.describe()\n",
    "# display(summary.T)\n",
    "# sns.scatterplot(data = summary.T, x = \"count\", y = \"max\")\n",
    "# plt.show()\n",
    "# sns.scatterplot(data = summary.T, x = \"count\", y = \"std\")\n",
    "# plt.show()\n",
    "\n",
    "\n",
    "train_df = trainDataDownSampled\n",
    "test_df = testData\n",
    "\n",
    "trainDataClass = train_df[\"Outcome\"]\n",
    "trainDataIds = train_df[\"Id\"]\n",
    "trainDataFeatures = train_df.drop(columns = [\"Outcome\", \"Id\"])\n",
    "\n",
    "testDataIds = test_df[\"Id\"]\n",
    "testDataFeatures = test_df.drop(columns = [\"Id\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "id": "f151620b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[Pipeline] ........... (step 1 of 3) Processing Scaling, total=   0.3s\n",
      "[Pipeline] .......... (step 2 of 3) Processing Imputing, total=  26.3s\n",
      "[Pipeline] ........ (step 3 of 3) Processing Regression, total=   2.4s\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>Id</th>\n",
       "      <th>Outcome</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>473</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>273</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1885</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>629</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>609</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>475</th>\n",
       "      <td>1182</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>476</th>\n",
       "      <td>978</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>477</th>\n",
       "      <td>1716</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>478</th>\n",
       "      <td>766</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>479</th>\n",
       "      <td>1155</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>480 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Id  Outcome\n",
       "0     473        2\n",
       "1     273        0\n",
       "2    1885        1\n",
       "3     629        3\n",
       "4     609        3\n",
       "..    ...      ...\n",
       "475  1182        1\n",
       "476   978        0\n",
       "477  1716        3\n",
       "478   766        2\n",
       "479  1155        3\n",
       "\n",
       "[480 rows x 2 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# exercise 3\n",
    "X_train, X_test, y_train, y_test, id_train, id_test = train_test_split(trainDataFeatures, trainDataClass, trainDataIds, test_size = .2, train_size = .8, random_state=24)\n",
    "\n",
    "e3_model = Pipeline([\n",
    "    (\"Scaling\", StandardScaler()), \n",
    "    (\"Imputing\", KNNImputer()), \n",
    "    (\"Regression\", LogisticRegression())], \n",
    "    verbose = True)\n",
    "\n",
    "e3_model.fit(X_train, y_train)\n",
    "\n",
    "# get model predictions on test set and output to .csv\n",
    "ids = pd.DataFrame(testDataIds, columns=[\"Id\"])\n",
    "e3_outcome = pd.DataFrame(e3_model.predict(testDataFeatures), columns=[\"Outcome\"])\n",
    "ids.index = range(len(e3_outcome))\n",
    "e3_outcome = pd.concat([ids, e3_outcome], axis=1)\n",
    "display(e3_outcome)\n",
    "e3_outcome.to_csv(\"e3_outcome.csv\", index=False)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8a4b16b6",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "8226bb59",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[Pipeline] ........... (step 1 of 2) Processing Scaling, total=   0.4s\n",
      "[Pipeline] .......... (step 2 of 2) Processing Imputing, total= 1.2min\n",
      "PC1 represents 0.41% of the variance.\n",
      "Gene 662 is the most contributing feature for PC1.\n",
      "\n",
      "PC2 represents 0.40% of the variance.\n",
      "Gene 1983 is the most contributing feature for PC2.\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.2s\n",
      "[Pipeline] .......... (step 2 of 4) Processing Imputing, total=  54.6s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.0s\n",
      "[Pipeline] ........ (step 4 of 4) Processing Regression, total=   0.1s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.2s\n",
      "[Pipeline] .......... (step 2 of 4) Processing Imputing, total=  53.1s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   0.7s\n",
      "[Pipeline] ........ (step 4 of 4) Processing Regression, total=   0.1s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.2s\n",
      "[Pipeline] .......... (step 2 of 4) Processing Imputing, total=  54.7s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   0.7s\n",
      "[Pipeline] ........ (step 4 of 4) Processing Regression, total=   0.0s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.2s\n",
      "[Pipeline] .......... (step 2 of 4) Processing Imputing, total=  51.7s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   0.7s\n",
      "[Pipeline] ........ (step 4 of 4) Processing Regression, total=   0.0s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.2s\n",
      "[Pipeline] .......... (step 2 of 4) Processing Imputing, total=  56.2s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   0.7s\n",
      "[Pipeline] ........ (step 4 of 4) Processing Regression, total=   0.1s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.2s\n",
      "[Pipeline] .......... (step 2 of 4) Processing Imputing, total=  60.0s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   0.7s\n",
      "[Pipeline] ........ (step 4 of 4) Processing Regression, total=   0.1s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.2s\n",
      "[Pipeline] .......... (step 2 of 4) Processing Imputing, total=  55.2s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   0.7s\n",
      "[Pipeline] ........ (step 4 of 4) Processing Regression, total=   0.1s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.2s\n",
      "[Pipeline] .......... (step 2 of 4) Processing Imputing, total=  52.4s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   0.7s\n",
      "[Pipeline] ........ (step 4 of 4) Processing Regression, total=   0.0s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.2s\n",
      "[Pipeline] .......... (step 2 of 4) Processing Imputing, total=  59.3s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   0.7s\n",
      "[Pipeline] ........ (step 4 of 4) Processing Regression, total=   0.1s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.2s\n",
      "[Pipeline] .......... (step 2 of 4) Processing Imputing, total= 1.0min\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   0.8s\n",
      "[Pipeline] ........ (step 4 of 4) Processing Regression, total=   0.1s\n",
      "The mean cross-validation accuracy is: 0.314 +/- 0.016\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.3s\n",
      "[Pipeline] .......... (step 2 of 4) Processing Imputing, total= 1.2min\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.0s\n",
      "[Pipeline] ........ (step 4 of 4) Processing Regression, total=   0.1s\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>Id</th>\n",
       "      <th>Outcome</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>473</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>273</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1885</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>629</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>609</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>475</th>\n",
       "      <td>1182</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>476</th>\n",
       "      <td>978</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>477</th>\n",
       "      <td>1716</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>478</th>\n",
       "      <td>766</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>479</th>\n",
       "      <td>1155</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>480 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Id  Outcome\n",
       "0     473        3\n",
       "1     273        3\n",
       "2    1885        3\n",
       "3     629        3\n",
       "4     609        3\n",
       "..    ...      ...\n",
       "475  1182        3\n",
       "476   978        3\n",
       "477  1716        3\n",
       "478   766        3\n",
       "479  1155        3\n",
       "\n",
       "[480 rows x 2 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# exercise 4\n",
    "\n",
    "# first just PCA\n",
    "# data normalization pipeline\n",
    "e4_normalized = Pipeline([\n",
    "    (\"Scaling\", StandardScaler()), \n",
    "    (\"Imputing\", KNNImputer())], \n",
    "    verbose = True)\n",
    "trainNormalized = e4_normalized.fit_transform(trainDataFeatures, trainDataClass)\n",
    "\n",
    "pcaModel = PCA(n_components = 2)\n",
    "e4_pca_data = pd.DataFrame(pcaModel.fit_transform(trainNormalized, trainDataClass), columns =[\"PC1\", \"PC2\"])\n",
    "pcs_var = pcaModel.explained_variance_ratio_\n",
    "for i in range(2):\n",
    "    print(f\"PC{i+1} represents {(pcs_var[i]*100):.2f}% of the variance.\")\n",
    "    mapping = (np.where(pcaModel.components_[i].max()==pcaModel.components_[i], True, False))\n",
    "    gene = (trainDataFeatures.columns[mapping])\n",
    "    print(f\"{gene[0]} is the most contributing feature for PC{i+1}.\\n\")\n",
    "sns.scatterplot(data = e4_pca_data, x=\"PC1\", y=\"PC2\")\n",
    "plt.show()\n",
    "\n",
    "# pipeline with PCA\n",
    "e4_model = Pipeline([\n",
    "    (\"Scaling\", StandardScaler()), \n",
    "    (\"Imputing\", KNNImputer()),\n",
    "    (\"PCA\", PCA(n_components = 200)),\n",
    "    (\"Regression\", LogisticRegression())], \n",
    "    verbose = True)\n",
    "\n",
    "cross_val = cross_validate(e4_model, trainDataFeatures, trainDataClass, cv = 10)\n",
    "\n",
    "scores = cross_val[\"test_score\"]\n",
    "print(f\"The mean cross-validation accuracy is: {scores.mean():.3f} +/- {scores.std():.3f}\")\n",
    "\n",
    "e4_model.fit(trainDataFeatures, trainDataClass)\n",
    "\n",
    "# get model predictions on test set and output to .csv\n",
    "ids = pd.DataFrame(testDataIds, columns=[\"Id\"])\n",
    "e4_outcome = pd.DataFrame(e4_model.predict(testDataFeatures), columns=[\"Outcome\"])\n",
    "ids.index = range(len(e4_outcome))\n",
    "e4_outcome = pd.concat([ids, e4_outcome], axis=1)\n",
    "display(e4_outcome)\n",
    "e4_outcome.to_csv(\"e4_outcome.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "f8804d1a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.4s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total=  59.7s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.1s\n",
      "[Pipeline] ..... (step 4 of 4) Processing RandomForrest, total=   3.0s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.2s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total= 1.2min\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   0.8s\n",
      "[Pipeline] ..... (step 4 of 4) Processing RandomForrest, total=   1.8s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.2s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total= 1.0min\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   0.8s\n",
      "[Pipeline] ..... (step 4 of 4) Processing RandomForrest, total=   1.9s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.3s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total= 1.3min\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.1s\n",
      "[Pipeline] ..... (step 4 of 4) Processing RandomForrest, total=   2.2s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   1.6s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total= 1.3min\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.0s\n",
      "[Pipeline] ..... (step 4 of 4) Processing RandomForrest, total=   2.1s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.3s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total= 1.1min\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.0s\n",
      "[Pipeline] ..... (step 4 of 4) Processing RandomForrest, total=   2.7s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.3s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total= 1.0min\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.6s\n",
      "[Pipeline] ..... (step 4 of 4) Processing RandomForrest, total=   2.1s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.3s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total= 1.1min\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.0s\n",
      "[Pipeline] ..... (step 4 of 4) Processing RandomForrest, total=   2.0s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.3s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total= 1.3min\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.0s\n",
      "[Pipeline] ..... (step 4 of 4) Processing RandomForrest, total=   2.0s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.3s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total= 1.4min\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   0.9s\n",
      "[Pipeline] ..... (step 4 of 4) Processing RandomForrest, total=   2.2s\n",
      "The mean cross-validation accuracy is: 0.377 +/- 0.003\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.4s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total= 1.4min\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.0s\n",
      "[Pipeline] ..... (step 4 of 4) Processing RandomForrest, total=   2.3s\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>Id</th>\n",
       "      <th>Outcome</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>473</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>273</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1885</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>629</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>609</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>475</th>\n",
       "      <td>1182</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>476</th>\n",
       "      <td>978</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>477</th>\n",
       "      <td>1716</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>478</th>\n",
       "      <td>766</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>479</th>\n",
       "      <td>1155</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>480 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Id  Outcome\n",
       "0     473        3\n",
       "1     273        3\n",
       "2    1885        3\n",
       "3     629        3\n",
       "4     609        3\n",
       "..    ...      ...\n",
       "475  1182        3\n",
       "476   978        3\n",
       "477  1716        3\n",
       "478   766        3\n",
       "479  1155        3\n",
       "\n",
       "[480 rows x 2 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# exercise 6\n",
    "e6_model = Pipeline([\n",
    "    (\"Scaling\", StandardScaler()),\n",
    "    (\"Imputation\", KNNImputer()),\n",
    "    (\"PCA\", PCA(n_components=200)),\n",
    "    (\"RandomForrest\", RandomForestClassifier())],\n",
    "    verbose=True)\n",
    "\n",
    "e6_cv_output = cross_validate(e6_model, trainDataFeatures, trainDataClass, cv = 10)\n",
    "\n",
    "scores = e6_cv_output[\"test_score\"]\n",
    "print(f\"The mean cross-validation accuracy is: {scores.mean():.3f} +/- {scores.std():.3f}\")\n",
    "\n",
    "e6_model.fit(trainDataFeatures, trainDataClass)\n",
    "\n",
    "# get model predictions on test set and output to .csv\n",
    "ids = pd.DataFrame(testDataIds, columns=[\"Id\"])\n",
    "e6_outcome = pd.DataFrame(e6_model.predict(testDataFeatures), columns=[\"Outcome\"])\n",
    "ids.index = range(len(e6_outcome))\n",
    "e6_outcome = pd.concat([ids, e6_outcome], axis=1)\n",
    "display(e6_outcome)\n",
    "e6_outcome.to_csv(\"e6_outcome.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "44295fd0",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "68c00315",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/v5/n_y2gqg51cgd0xj05l2c_c7h0000gn/T/ipykernel_6592/1685860185.py:19: DeprecationWarning: KerasClassifier is deprecated, use Sci-Keras (https://github.com/adriangb/scikeras) instead.\n",
      "  (\"DenseNet\", KerasClassifier(build_fn=neural_net, verbose=0))],\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.8s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total=   1.1s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   2.0s\n",
      "[Pipeline] .......... (step 4 of 4) Processing DenseNet, total=   2.1s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.3s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total=   0.3s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.2s\n",
      "[Pipeline] .......... (step 4 of 4) Processing DenseNet, total=   1.4s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.3s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total=   0.3s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.2s\n",
      "[Pipeline] .......... (step 4 of 4) Processing DenseNet, total=   1.4s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.3s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total=   0.2s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.8s\n",
      "[Pipeline] .......... (step 4 of 4) Processing DenseNet, total=   3.1s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.3s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total=   0.2s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.2s\n",
      "[Pipeline] .......... (step 4 of 4) Processing DenseNet, total=   1.4s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.3s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total=   0.2s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.2s\n",
      "[Pipeline] .......... (step 4 of 4) Processing DenseNet, total=   1.4s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.3s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total=   0.2s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.3s\n",
      "[Pipeline] .......... (step 4 of 4) Processing DenseNet, total=   2.0s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.5s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total=   0.4s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.4s\n",
      "[Pipeline] .......... (step 4 of 4) Processing DenseNet, total=   1.6s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.3s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total=   0.2s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.3s\n",
      "[Pipeline] .......... (step 4 of 4) Processing DenseNet, total=   2.3s\n",
      "[Pipeline] ........... (step 1 of 4) Processing Scaling, total=   0.3s\n",
      "[Pipeline] ........ (step 2 of 4) Processing Imputation, total=   0.2s\n",
      "[Pipeline] ............... (step 3 of 4) Processing PCA, total=   1.7s\n",
      "[Pipeline] .......... (step 4 of 4) Processing DenseNet, total=   1.4s\n"
     ]
    }
   ],
   "source": [
    "# exercise 8\n",
    "\n",
    "def neural_net():\n",
    "    model = Sequential([\n",
    "        Dense(30, activation=\"relu\"),\n",
    "        Dense(20, activation=\"relu\"),\n",
    "        Dense(10, activation=\"relu\"),\n",
    "        Dense(4, activation=keras.activations.softmax)\n",
    "    ])\n",
    "    model.compile(\n",
    "        loss=\"sparse_categorical_crossentropy\",\n",
    "        metrics=[\"accuracy\"])\n",
    "    return model\n",
    "\n",
    "dense_net_pipe = Pipeline([\n",
    "    (\"Scaling\", StandardScaler()),\n",
    "    (\"Imputation\", KNNImputer()),\n",
    "    (\"PCA\", PCA(n_components=200)),\n",
    "    (\"DenseNet\", KerasClassifier(build_fn=neural_net, verbose=0))],\n",
    "    verbose=True)\n",
    "\n",
    "e8_cv_output = cross_validate(dense_net_pipe, trainDataFeatures, trainDataClass, cv=10, return_train_score=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "1c1153be",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'fit_time': array([6.28660178, 3.28534293, 3.24922299, 5.49069905, 3.0823729 ,\n",
       "        3.07962298, 3.91343832, 3.97932625, 4.17541814, 3.58819199]),\n",
       " 'score_time': array([0.74127531, 0.47551107, 0.41399503, 0.42026424, 0.42754102,\n",
       "        0.42734599, 0.48678493, 0.46319175, 0.41339183, 0.52518606]),\n",
       " 'test_score': array([0.21710527, 0.29605263, 0.32894737, 0.23026316, 0.2631579 ,\n",
       "        0.22368421, 0.24342105, 0.27631578, 0.30921054, 0.32894737]),\n",
       " 'train_score': array([0.36038011, 0.36257309, 0.40350878, 0.37061402, 0.37646198,\n",
       "        0.34429824, 0.3128655 , 0.36622807, 0.40204677, 0.36622807])}"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The mean cross-validation accuracy is: 0.272 +/- 0.041\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "\"it is 'so much' better\""
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "display(e8_cv_output)\n",
    "scores = e8_cv_output[\"test_score\"]\n",
    "print(f\"The mean cross-validation accuracy is: {scores.mean():.3f} +/- {scores.std():.3f}\")\n",
    "\"\"\"it is 'so much' better\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "85add1f1",
   "metadata": {},
   "outputs": [],
   "source": [
    "e8_model.fit(trainDataFeatures, trainDataClass)\n",
    "\n",
    "# get model predictions on test set and output to .csv\n",
    "ids = pd.DataFrame(testDataIds, columns=[\"Id\"])\n",
    "e8_outcome = pd.DataFrame(e_model.predict(testDataFeatures), columns=[\"Outcome\"])\n",
    "ids.index = range(len(e6_outcome))\n",
    "e6_outcome = pd.concat([ids, e6_outcome], axis=1)\n",
    "display(e6_outcome)\n",
    "e6_outcome.to_csv(\"e6_outcome.csv\", index=False)"
   ]
  }
 ],
 "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}