4-ML/1-process_data.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "toc": true
   },
   "source": [
    "<h1>Table of Contents<span class=\"tocSkip\"></span></h1>\n",
    "<div class=\"toc\"><ul class=\"toc-item\"><li><span><a href=\"#Data-processing-for-pet-detection\" data-toc-modified-id=\"Data-processing-for-pet-detection-1\"><span class=\"toc-item-num\">1&nbsp;&nbsp;</span>Data processing for pet detection</a></span></li><li><span><a href=\"#Read-the-data\" data-toc-modified-id=\"Read-the-data-2\"><span class=\"toc-item-num\">2&nbsp;&nbsp;</span>Read the data</a></span></li><li><span><a href=\"#ML-Part\" data-toc-modified-id=\"ML-Part-3\"><span class=\"toc-item-num\">3&nbsp;&nbsp;</span>ML Part</a></span><ul class=\"toc-item\"><li><span><a href=\"#ML---full-evaluation\" data-toc-modified-id=\"ML---full-evaluation-3.1\"><span class=\"toc-item-num\">3.1&nbsp;&nbsp;</span>ML - full evaluation</a></span></li><li><span><a href=\"#Let's-probe-preprocessing-methods-with-MLP\" data-toc-modified-id=\"Let's-probe-preprocessing-methods-with-MLP-3.2\"><span class=\"toc-item-num\">3.2&nbsp;&nbsp;</span>Let's probe preprocessing methods with MLP</a></span><ul class=\"toc-item\"><li><span><a href=\"#Conclusions\" data-toc-modified-id=\"Conclusions-3.2.1\"><span class=\"toc-item-num\">3.2.1&nbsp;&nbsp;</span>Conclusions</a></span></li></ul></li></ul></li><li><span><a href=\"#Let's-build-the-models-with-the-best-ML-methods\" data-toc-modified-id=\"Let's-build-the-models-with-the-best-ML-methods-4\"><span class=\"toc-item-num\">4&nbsp;&nbsp;</span>Let's build the models with the best ML methods</a></span></li><li><span><a href=\"#Save-models-for-future-use\" data-toc-modified-id=\"Save-models-for-future-use-5\"><span class=\"toc-item-num\">5&nbsp;&nbsp;</span>Save models for future use</a></span></li><li><span><a href=\"#Reuse-saved-models-for-new-predictions\" data-toc-modified-id=\"Reuse-saved-models-for-new-predictions-6\"><span class=\"toc-item-num\">6&nbsp;&nbsp;</span>Reuse saved models for new predictions</a></span></li></ul></div>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-05T14:56:43.841320Z",
     "start_time": "2021-01-05T14:56:43.838132Z"
    }
   },
   "source": [
    "# Data processing for pet detection\n",
    "\n",
    "Here I'm preparing database csv dump for use with machine learning, then probing some ML algorithms and tuning them, and finally building models with best selected algorithm."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T17:41:17.561084Z",
     "start_time": "2021-01-18T17:41:17.553784Z"
    }
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "plt.rcParams['figure.figsize'] = [10, 10]\n",
    "\n",
    "\n",
    "import seaborn as sns\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "%matplotlib inline\n",
    "\n",
    "import datetime"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T17:41:17.575579Z",
     "start_time": "2021-01-18T17:41:17.563439Z"
    }
   },
   "outputs": [],
   "source": [
    "# Let's ignore the warnings\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T17:41:17.588422Z",
     "start_time": "2021-01-18T17:41:17.578991Z"
    }
   },
   "outputs": [],
   "source": [
    "# ML modules\n",
    "\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.neighbors import RadiusNeighborsClassifier\n",
    "\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.preprocessing import MaxAbsScaler\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.preprocessing import RobustScaler\n",
    "from sklearn.preprocessing import Normalizer\n",
    "from sklearn.preprocessing import QuantileTransformer\n",
    "from sklearn.preprocessing import PowerTransformer\n",
    "from sklearn.preprocessing import KBinsDiscretizer\n",
    "from sklearn.preprocessing import PolynomialFeatures\n",
    "\n",
    "from sklearn.preprocessing import FunctionTransformer\n",
    "from sklearn.pipeline import make_pipeline\n",
    "\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import ShuffleSplit\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "from sklearn import metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T17:41:17.597837Z",
     "start_time": "2021-01-18T17:41:17.591401Z"
    }
   },
   "outputs": [],
   "source": [
    "sns.set(font=\"monospace\")\n",
    "sns.set_style(\"white\")\n",
    "sns.set(color_codes=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Read the data\n",
    "\n",
    "- read the input data\n",
    "- process it (pivot) to the form acceptable by the ml methods"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T17:41:17.721179Z",
     "start_time": "2021-01-18T17:41:17.600151Z"
    }
   },
   "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>_value</th>\n",
       "      <th>gdzie</th>\n",
       "      <th>kto</th>\n",
       "      <th>location</th>\n",
       "      <th>source</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>target</th>\n",
       "      <th>timegroup</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>index</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-42.0</td>\n",
       "      <td>dol</td>\n",
       "      <td>miko</td>\n",
       "      <td>kabaty</td>\n",
       "      <td>ttgo1</td>\n",
       "      <td>2021-01-03 18:07:19.000003600</td>\n",
       "      <td>bathroom_d</td>\n",
       "      <td>2021-01-03 18:07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-59.0</td>\n",
       "      <td>dol</td>\n",
       "      <td>miko</td>\n",
       "      <td>kabaty</td>\n",
       "      <td>ttgo1</td>\n",
       "      <td>2021-01-03 18:08:19.000003600</td>\n",
       "      <td>bathroom_d</td>\n",
       "      <td>2021-01-03 18:08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-63.0</td>\n",
       "      <td>dol</td>\n",
       "      <td>miko</td>\n",
       "      <td>kabaty</td>\n",
       "      <td>ttgo1</td>\n",
       "      <td>2021-01-03 18:09:19.000003600</td>\n",
       "      <td>bathroom_d</td>\n",
       "      <td>2021-01-03 18:09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-63.0</td>\n",
       "      <td>dol</td>\n",
       "      <td>miko</td>\n",
       "      <td>kabaty</td>\n",
       "      <td>ttgo1</td>\n",
       "      <td>2021-01-03 18:10:19.000003600</td>\n",
       "      <td>bathroom_d</td>\n",
       "      <td>2021-01-03 18:10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-66.0</td>\n",
       "      <td>dol</td>\n",
       "      <td>miko</td>\n",
       "      <td>kabaty</td>\n",
       "      <td>ttgo1</td>\n",
       "      <td>2021-01-03 18:11:19.000003600</td>\n",
       "      <td>bathroom_d</td>\n",
       "      <td>2021-01-03 18:11</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>1920</th>\n",
       "      <td>-84.0</td>\n",
       "      <td>strych</td>\n",
       "      <td>miko</td>\n",
       "      <td>kabaty</td>\n",
       "      <td>esp32-strych</td>\n",
       "      <td>2021-01-03 16:38:16.000003600</td>\n",
       "      <td>room_m</td>\n",
       "      <td>2021-01-03 16:38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1921</th>\n",
       "      <td>-83.0</td>\n",
       "      <td>strych</td>\n",
       "      <td>miko</td>\n",
       "      <td>kabaty</td>\n",
       "      <td>esp32-strych</td>\n",
       "      <td>2021-01-03 16:39:16.000003600</td>\n",
       "      <td>room_m</td>\n",
       "      <td>2021-01-03 16:39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1922</th>\n",
       "      <td>-67.0</td>\n",
       "      <td>strych</td>\n",
       "      <td>miko</td>\n",
       "      <td>kabaty</td>\n",
       "      <td>esp32-strych</td>\n",
       "      <td>2021-01-03 16:40:17.000003600</td>\n",
       "      <td>room_m</td>\n",
       "      <td>2021-01-03 16:40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1923</th>\n",
       "      <td>-81.0</td>\n",
       "      <td>strych</td>\n",
       "      <td>miko</td>\n",
       "      <td>kabaty</td>\n",
       "      <td>esp32-strych</td>\n",
       "      <td>2021-01-03 16:41:17.000003600</td>\n",
       "      <td>room_m</td>\n",
       "      <td>2021-01-03 16:41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1924</th>\n",
       "      <td>-82.0</td>\n",
       "      <td>strych</td>\n",
       "      <td>miko</td>\n",
       "      <td>kabaty</td>\n",
       "      <td>esp32-strych</td>\n",
       "      <td>2021-01-03 16:42:17.000003600</td>\n",
       "      <td>room_m</td>\n",
       "      <td>2021-01-03 16:42</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1925 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       _value   gdzie   kto location        source  \\\n",
       "index                                                \n",
       "0       -42.0     dol  miko   kabaty         ttgo1   \n",
       "1       -59.0     dol  miko   kabaty         ttgo1   \n",
       "2       -63.0     dol  miko   kabaty         ttgo1   \n",
       "3       -63.0     dol  miko   kabaty         ttgo1   \n",
       "4       -66.0     dol  miko   kabaty         ttgo1   \n",
       "...       ...     ...   ...      ...           ...   \n",
       "1920    -84.0  strych  miko   kabaty  esp32-strych   \n",
       "1921    -83.0  strych  miko   kabaty  esp32-strych   \n",
       "1922    -67.0  strych  miko   kabaty  esp32-strych   \n",
       "1923    -81.0  strych  miko   kabaty  esp32-strych   \n",
       "1924    -82.0  strych  miko   kabaty  esp32-strych   \n",
       "\n",
       "                          timestamp      target         timegroup  \n",
       "index                                                              \n",
       "0     2021-01-03 18:07:19.000003600  bathroom_d  2021-01-03 18:07  \n",
       "1     2021-01-03 18:08:19.000003600  bathroom_d  2021-01-03 18:08  \n",
       "2     2021-01-03 18:09:19.000003600  bathroom_d  2021-01-03 18:09  \n",
       "3     2021-01-03 18:10:19.000003600  bathroom_d  2021-01-03 18:10  \n",
       "4     2021-01-03 18:11:19.000003600  bathroom_d  2021-01-03 18:11  \n",
       "...                             ...         ...               ...  \n",
       "1920  2021-01-03 16:38:16.000003600      room_m  2021-01-03 16:38  \n",
       "1921  2021-01-03 16:39:16.000003600      room_m  2021-01-03 16:39  \n",
       "1922  2021-01-03 16:40:17.000003600      room_m  2021-01-03 16:40  \n",
       "1923  2021-01-03 16:41:17.000003600      room_m  2021-01-03 16:41  \n",
       "1924  2021-01-03 16:42:17.000003600      room_m  2021-01-03 16:42  \n",
       "\n",
       "[1925 rows x 8 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# read and handle data\n",
    "\n",
    "fulldata = pd.read_csv(\"1-data/all-20210103.csv\", delimiter=\";\")  # header=None\n",
    "fulldata.drop(['ignore'], axis=1, inplace=True)\n",
    "fulldata.reset_index()\n",
    "fulldata.index.name = 'index'\n",
    "\n",
    "\n",
    "# time to time column\n",
    "fulldata['timestamp'] = pd.to_datetime(fulldata['timestamp'], format='%Y-%m-%d %H:%M:%S.%f')\n",
    "\n",
    "fulldata['timegroup'] = fulldata.timestamp.dt.strftime('%Y-%m-%d %H:%M')\n",
    "\n",
    "fulldata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T17:41:17.786166Z",
     "start_time": "2021-01-18T17:41:17.724309Z"
    }
   },
   "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>source</th>\n",
       "      <th>timegroup</th>\n",
       "      <th>target</th>\n",
       "      <th>esp32-strych</th>\n",
       "      <th>esp32-tm1637</th>\n",
       "      <th>esp32cam-1</th>\n",
       "      <th>ttgo1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2021-01-03 15:32</td>\n",
       "      <td>office</td>\n",
       "      <td>-71.0</td>\n",
       "      <td>-110.0</td>\n",
       "      <td>-71.5</td>\n",
       "      <td>-83.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2021-01-03 15:33</td>\n",
       "      <td>office</td>\n",
       "      <td>-73.0</td>\n",
       "      <td>-110.0</td>\n",
       "      <td>-82.5</td>\n",
       "      <td>-82.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2021-01-03 15:34</td>\n",
       "      <td>office</td>\n",
       "      <td>-81.0</td>\n",
       "      <td>-110.0</td>\n",
       "      <td>-80.0</td>\n",
       "      <td>-87.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2021-01-03 15:35</td>\n",
       "      <td>office</td>\n",
       "      <td>-75.0</td>\n",
       "      <td>-110.0</td>\n",
       "      <td>-86.5</td>\n",
       "      <td>-86.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2021-01-03 15:36</td>\n",
       "      <td>office</td>\n",
       "      <td>-72.5</td>\n",
       "      <td>-110.0</td>\n",
       "      <td>-78.5</td>\n",
       "      <td>-81.0</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>236</th>\n",
       "      <td>2021-01-03 21:56</td>\n",
       "      <td>room_k</td>\n",
       "      <td>-100.5</td>\n",
       "      <td>-67.5</td>\n",
       "      <td>-110.0</td>\n",
       "      <td>-91.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>237</th>\n",
       "      <td>2021-01-03 21:57</td>\n",
       "      <td>room_k</td>\n",
       "      <td>-86.0</td>\n",
       "      <td>-60.0</td>\n",
       "      <td>-110.0</td>\n",
       "      <td>-91.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>238</th>\n",
       "      <td>2021-01-03 21:58</td>\n",
       "      <td>room_k</td>\n",
       "      <td>-88.5</td>\n",
       "      <td>-93.5</td>\n",
       "      <td>-110.0</td>\n",
       "      <td>-91.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>239</th>\n",
       "      <td>2021-01-03 21:59</td>\n",
       "      <td>room_k</td>\n",
       "      <td>-86.0</td>\n",
       "      <td>-75.0</td>\n",
       "      <td>-110.0</td>\n",
       "      <td>-91.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>240</th>\n",
       "      <td>2021-01-03 22:00</td>\n",
       "      <td>room_k</td>\n",
       "      <td>-87.0</td>\n",
       "      <td>-72.0</td>\n",
       "      <td>-110.0</td>\n",
       "      <td>-84.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>240 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "source         timegroup  target  esp32-strych  esp32-tm1637  esp32cam-1  \\\n",
       "0       2021-01-03 15:32  office         -71.0        -110.0       -71.5   \n",
       "1       2021-01-03 15:33  office         -73.0        -110.0       -82.5   \n",
       "2       2021-01-03 15:34  office         -81.0        -110.0       -80.0   \n",
       "3       2021-01-03 15:35  office         -75.0        -110.0       -86.5   \n",
       "4       2021-01-03 15:36  office         -72.5        -110.0       -78.5   \n",
       "..                   ...     ...           ...           ...         ...   \n",
       "236     2021-01-03 21:56  room_k        -100.5         -67.5      -110.0   \n",
       "237     2021-01-03 21:57  room_k         -86.0         -60.0      -110.0   \n",
       "238     2021-01-03 21:58  room_k         -88.5         -93.5      -110.0   \n",
       "239     2021-01-03 21:59  room_k         -86.0         -75.0      -110.0   \n",
       "240     2021-01-03 22:00  room_k         -87.0         -72.0      -110.0   \n",
       "\n",
       "source  ttgo1  \n",
       "0       -83.0  \n",
       "1       -82.5  \n",
       "2       -87.5  \n",
       "3       -86.5  \n",
       "4       -81.0  \n",
       "..        ...  \n",
       "236     -91.5  \n",
       "237     -91.5  \n",
       "238     -91.5  \n",
       "239     -91.5  \n",
       "240     -84.5  \n",
       "\n",
       "[240 rows x 6 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# pivot the data\n",
    "\n",
    "fulldataPivot = pd.pivot_table(fulldata,\n",
    "                               index=['timegroup', 'target'], columns='source', values='_value', margins=False)\n",
    "\n",
    "fulldataPivot = fulldataPivot.reset_index().dropna(axis='rows') #.drop(['timegroup', 'source'], axis=1, inplace=True)\n",
    "\n",
    "fulldataPivot.to_csv(\"1-data/all-20210103-pivot.csv\", sep=\";\")\n",
    "\n",
    "fulldataPivot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-13T18:53:05.264468Z",
     "start_time": "2021-01-13T18:53:05.253947Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "target\n",
       "bathroom_d      11\n",
       "bathroom_g      30\n",
       "living_room    101\n",
       "office          38\n",
       "room_k          32\n",
       "room_m          28\n",
       "Name: timegroup, dtype: int64"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# number of data points in different rooms\n",
    "\n",
    "fulldataPivot.groupby(\"target\").count()[\"timegroup\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ML Part"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T17:41:56.005649Z",
     "start_time": "2021-01-18T17:41:56.001778Z"
    }
   },
   "outputs": [],
   "source": [
    "# split data to signals and class (target)\n",
    "\n",
    "y = fulldataPivot['target']\n",
    "X = fulldataPivot[['esp32-strych', 'esp32-tm1637', 'esp32cam-1', 'ttgo1']]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## ML - full evaluation\n",
    "\n",
    "Let's screen some algorithms. We will test both: ML algorithms and data preprocessing (such as normalization). Next we will optimize the best performing method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-09T19:17:32.032893Z",
     "start_time": "2021-01-09T19:17:31.996105Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NoScaling\n",
      "             RadiusNeighborsClassifier\n",
      "             DecisionTreeClassifier\n",
      "             KNeighborsClassifier\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      "KBinsDiscretizer-ordinal\n",
      "             RadiusNeighborsClassifier\n",
      "             DecisionTreeClassifier\n",
      "             KNeighborsClassifier\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      "KBinsDiscretizer-onehot\n",
      "             RadiusNeighborsClassifier\n",
      "             DecisionTreeClassifier\n",
      "             KNeighborsClassifier\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      "PolynomialFeatures-2\n",
      "             RadiusNeighborsClassifier\n",
      "             DecisionTreeClassifier\n",
      "             KNeighborsClassifier\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      "PolynomialFeatures-interaction_only\n",
      "             RadiusNeighborsClassifier\n",
      "             DecisionTreeClassifier\n",
      "             KNeighborsClassifier\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      "MinMaxScaler\n",
      "             RadiusNeighborsClassifier\n",
      "             DecisionTreeClassifier\n",
      "             KNeighborsClassifier\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      "PowerTransformer-yeo-johnson\n",
      "             RadiusNeighborsClassifier\n",
      "             DecisionTreeClassifier\n",
      "             KNeighborsClassifier\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      "StandardScaler\n",
      "             RadiusNeighborsClassifier\n",
      "             DecisionTreeClassifier\n",
      "             KNeighborsClassifier\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      "MaxAbsScaler\n",
      "             RadiusNeighborsClassifier\n",
      "             DecisionTreeClassifier\n",
      "             KNeighborsClassifier\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      "Normalizer\n",
      "             RadiusNeighborsClassifier\n",
      "             DecisionTreeClassifier\n",
      "             KNeighborsClassifier\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      "QuantileTransformer-normal\n",
      "             RadiusNeighborsClassifier\n",
      "             DecisionTreeClassifier\n",
      "             KNeighborsClassifier\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n"
     ]
    }
   ],
   "source": [
    "# Here we will construct combinations of ml methods and scalers.\n",
    "# Please reference https://scikit-learn.org/stable/modules/multiclass.html for algorithms\n",
    "# capable of handling multiclass.\n",
    "\n",
    "### Scalers\n",
    "\n",
    "dummyScaller = FunctionTransformer()\n",
    "\n",
    "scalers = [\n",
    "    (dummyScaller, \"NoScaling\"),\n",
    "    (KBinsDiscretizer(encode='ordinal', n_bins=6), 'KBinsDiscretizer-ordinal'),\n",
    "    (KBinsDiscretizer(encode='onehot', n_bins=6), 'KBinsDiscretizer-onehot'),\n",
    "    (PolynomialFeatures(2),'PolynomialFeatures-2'),\n",
    "    (PolynomialFeatures(interaction_only=True),'PolynomialFeatures-interaction_only'),\n",
    "    (MinMaxScaler(), \"MinMaxScaler\"),\n",
    "    (PowerTransformer(method='yeo-johnson'), \"PowerTransformer-yeo-johnson\"),\n",
    "    (StandardScaler(), \"StandardScaler\"),\n",
    "    (MaxAbsScaler(), \"MaxAbsScaler\"),\n",
    "#     (RobustScaler(quantile_range=(25, 75)), \"RobustScaler\"),\n",
    "    (Normalizer(), \"Normalizer\"),\n",
    "    (QuantileTransformer(output_distribution='normal'), \"QuantileTransformer-normal\"),\n",
    "#     (QuantileTransformer(output_distribution='uniform'), \"QuantileTransformer-uniform\"),\n",
    "#     (PowerTransformer(method='box-cox'), \"PowerTransformer box-cox\"),\n",
    "    \n",
    "]\n",
    "\n",
    "\n",
    "### classifiers\n",
    "\n",
    "classifiers = [\n",
    "    RadiusNeighborsClassifier(),\n",
    "    DecisionTreeClassifier(),\n",
    "    KNeighborsClassifier(),\n",
    "    RandomForestClassifier(),\n",
    "    MLPClassifier(max_iter=2048,),\n",
    "]\n",
    "\n",
    "\n",
    "### classifiers search space\n",
    "\n",
    "\n",
    "searchSpaceDict = {}\n",
    "\n",
    "searchSpaceDict[\"RandomForestClassifier\"] = {'randomforestclassifier__n_estimators': [20,50, 100, 120,150, 175, 200, 250, 300],\n",
    "                                            'randomforestclassifier__max_depth': [None, 5,10,20,30,40,50,80,100,120,200,220,250,300],\n",
    "                                           'randomforestclassifier__max_features': [\"auto\", \"sqrt\", \"log2\"],}\n",
    "\n",
    "searchSpaceDict[\"KNeighborsClassifier\"] = {'kneighborsclassifier__weights': ['uniform', 'distance'],\n",
    "                                          'kneighborsclassifier__n_neighbors': range(1,30),\n",
    "                                          'kneighborsclassifier__leaf_size': range(10,50)}\n",
    "\n",
    "\n",
    "searchSpaceDict[\"MLPClassifier\"] = {\n",
    "                            'mlpclassifier__hidden_layer_sizes': [(20,20), (50,50), (50,), (100,),(150,)],\n",
    "                            'mlpclassifier__activation': ['identity', 'logistic', 'tanh', 'relu'],\n",
    "                            'mlpclassifier__solver': ['lbfgs', 'sgd', 'adam'], }\n",
    "\n",
    "searchSpaceDict[\"DecisionTreeClassifier\"] = {}\n",
    "\n",
    "searchSpaceDict[\"RadiusNeighborsClassifier\"] = {'radiusneighborsclassifier__radius': range(1,3), \n",
    "                                               'radiusneighborsclassifier__weights': ['uniform', 'distance'],\n",
    "                                                'radiusneighborsclassifier__leaf_size': range(10,50)}\n",
    "\n",
    "\n",
    "## make the pipelines\n",
    "\n",
    "pipelinesMatrix = {}\n",
    "\n",
    "for scaler, scalerDesc in scalers:\n",
    "    pipelinesMatrix[scalerDesc] = {}\n",
    "    print(scalerDesc) \n",
    "    for classifier in classifiers:\n",
    "        print(\"            \", classifier.__class__.__name__)\n",
    "        pipelinesMatrix[scalerDesc][classifier.__class__.__name__] = make_pipeline(scaler, classifier)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-09T19:17:32.051578Z",
     "start_time": "2021-01-09T19:17:32.039271Z"
    }
   },
   "outputs": [],
   "source": [
    "# function for performing cross validation of all algorithms\n",
    "\n",
    "def cvAll(X, y, pipelinesMatrix):\n",
    "    i=0\n",
    "    for scaler in pipelinesMatrix:\n",
    "        print(\"----------------------\", scaler)\n",
    "        for classifier in pipelinesMatrix[scaler]:\n",
    "            i+=1\n",
    "            print(\"     +++++++\", classifier)\n",
    "            startT = datetime.datetime.now()\n",
    "\n",
    "            pipeline = pipelinesMatrix[scaler][classifier]\n",
    "\n",
    "                \n",
    "            searchSpace = searchSpaceDict[classifier]            \n",
    "            clf = GridSearchCV(pipeline,\n",
    "                                 searchSpace, scoring='balanced_accuracy',\n",
    "                                 n_jobs=4)\n",
    "\n",
    "            cv = ShuffleSplit(n_splits=10, test_size=0.333)\n",
    "            scores = cross_val_score(clf, X, y, cv=cv, scoring='balanced_accuracy')\n",
    "\n",
    "            print (\"          scores: \", scores)\n",
    "\n",
    "           \n",
    "            headers = ['scaler', 'classifier', 'balanced_accuracy', 'balanced_accuracy STD']\n",
    "            dfResultsTemp = pd.DataFrame(columns = headers)\n",
    "            dfResultsTemp.loc[0] = [scaler, classifier, scores.mean(), scores.std()]\n",
    "            \n",
    "            print(\"             exec time:\", datetime.datetime.now()-startT, datetime.datetime.now())\n",
    "\n",
    "            if i==1:\n",
    "                DataConcat = dfResultsTemp.copy()\n",
    "            else:\n",
    "                DataConcat = pd.concat([DataConcat, dfResultsTemp])\n",
    "\n",
    "    return DataConcat    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-10T07:49:10.997626Z",
     "start_time": "2021-01-09T19:17:32.060052Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---------------------- NoScaling\n",
      "     +++++++ RadiusNeighborsClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:00:14.637400 2021-01-09 20:17:46.699067\n",
      "     +++++++ DecisionTreeClassifier\n",
      "          scores:  [0.65291005 0.73662945 0.68194444 0.83681319 0.65823413 0.73571429\n",
      " 0.69447115 0.86064815 0.7712963  0.72638186]\n",
      "             exec time: 0:00:01.610719 2021-01-09 20:17:48.310409\n",
      "     +++++++ KNeighborsClassifier\n",
      "          scores:  [0.7519943  0.76525303 0.71664895 0.73373016 0.72859848 0.81443834\n",
      " 0.75870871 0.72197941 0.75188863 0.80490388]\n",
      "             exec time: 0:06:24.850912 2021-01-09 20:24:13.163436\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.86166761 0.88484848 0.75696248 0.75277778 0.83554293 0.77165992\n",
      " 0.72402159 0.7972941  0.81882932 0.72128427]\n",
      "             exec time: 0:35:44.885311 2021-01-09 20:59:58.050396\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.82859848 0.81022386 0.16666667 0.83310557 0.66116597 0.84502696\n",
      " 0.77333822 0.7156746  0.80555556 0.90636973]\n",
      "             exec time: 0:30:07.712796 2021-01-09 21:30:05.764891\n",
      "---------------------- KBinsDiscretizer-ordinal\n",
      "     +++++++ RadiusNeighborsClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:00:14.347165 2021-01-09 21:30:20.114216\n",
      "     +++++++ DecisionTreeClassifier\n",
      "          scores:  [0.66616162 0.58956395 0.67577625 0.82439782 0.67622073 0.69403236\n",
      " 0.61971691 0.73715225 0.6695288  0.6709964 ]\n",
      "             exec time: 0:00:02.033429 2021-01-09 21:30:22.149986\n",
      "     +++++++ KNeighborsClassifier\n",
      "          scores:  [0.70006556 0.64861111 0.7459596  0.74890605 0.70729662 0.80781025\n",
      " 0.76071429 0.63205128 0.72696927 0.6961527 ]\n",
      "             exec time: 0:09:45.062934 2021-01-09 21:40:07.214773\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.68186186 0.78040016 0.79265873 0.69529915 0.73131868 0.64279666\n",
      " 0.72799355 0.64720077 0.73928571 0.8001638 ]\n",
      "             exec time: 0:36:40.301489 2021-01-09 22:16:47.518489\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.68656209 0.7512617  0.71362434 0.8028083  0.73392857 0.72333822\n",
      " 0.67391659 0.81739418 0.80952381 0.7881734 ]\n",
      "             exec time: 0:25:24.577057 2021-01-09 22:42:12.096455\n",
      "---------------------- KBinsDiscretizer-onehot\n",
      "     +++++++ RadiusNeighborsClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:00:12.369794 2021-01-09 22:42:24.467120\n",
      "     +++++++ DecisionTreeClassifier\n",
      "          scores:  [0.78055556 0.59528897 0.61280569 0.73213083 0.71708754 0.76070076\n",
      " 0.67553763 0.63961761 0.60555556 0.73219373]\n",
      "             exec time: 0:00:02.097112 2021-01-09 22:42:26.566702\n",
      "     +++++++ KNeighborsClassifier\n",
      "          scores:  [0.68113426 0.59080474 0.61895161 0.69307359 0.71385049 0.72712648\n",
      " 0.71179467 0.65039683 0.63850289 0.53917379]\n",
      "             exec time: 0:11:42.267951 2021-01-09 22:54:08.836823\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.69934641 0.63941799 0.66132479 0.57273557 0.61845217 0.71894587\n",
      " 0.68426118 0.6649531  0.78564601 0.69879079]\n",
      "             exec time: 0:36:51.006203 2021-01-09 23:30:59.843942\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.74771242 0.70035885 0.79498564 0.80840456 0.72799423 0.70968302\n",
      " 0.75958012 0.65994916 0.68689274 0.7787037 ]\n",
      "             exec time: 0:25:35.925244 2021-01-09 23:56:35.771108\n",
      "---------------------- PolynomialFeatures-2\n",
      "     +++++++ RadiusNeighborsClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:00:12.053688 2021-01-09 23:56:47.826736\n",
      "     +++++++ DecisionTreeClassifier\n",
      "          scores:  [0.68769433 0.71582492 0.62600733 0.68889308 0.74145759 0.69493177\n",
      " 0.84401946 0.67514881 0.57808858 0.70254047]\n",
      "             exec time: 0:00:01.744603 2021-01-09 23:56:49.573085\n",
      "     +++++++ KNeighborsClassifier\n",
      "          scores:  [0.8988054  0.84594767 0.78366883 0.7146895  0.78818911 0.76353276\n",
      " 0.73332635 0.66316739 0.8107906  0.77955748]\n",
      "             exec time: 0:07:53.810242 2021-01-10 00:04:43.384236\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.74102564 0.61536797 0.81267696 0.81861472 0.68571429 0.66626984\n",
      " 0.80495762 0.80164093 0.77619949 0.79166231]\n",
      "             exec time: 0:36:01.078873 2021-01-10 00:40:44.464009\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.64493064 0.65829772 0.75980071 0.81709402 0.64586571 0.70970418\n",
      " 0.66810501 0.67063492 0.63472823 0.81304563]\n",
      "             exec time: 0:14:12.486950 2021-01-10 00:54:56.952764\n",
      "---------------------- PolynomialFeatures-interaction_only\n",
      "     +++++++ RadiusNeighborsClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:00:12.065603 2021-01-10 00:55:09.019842\n",
      "     +++++++ DecisionTreeClassifier\n",
      "          scores:  [0.62051282 0.76342084 0.6849757  0.73178105 0.66111111 0.77637918\n",
      " 0.81575092 0.71694398 0.65307359 0.75889111]\n",
      "             exec time: 0:00:01.709844 2021-01-10 00:55:10.731002\n",
      "     +++++++ KNeighborsClassifier\n",
      "          scores:  [0.73749399 0.72539683 0.79313023 0.8107684  0.77519182 0.80598846\n",
      " 0.76479076 0.68875063 0.80927871 0.70073529]\n",
      "             exec time: 0:07:51.466237 2021-01-10 01:03:02.198864\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.80273199 0.72853535 0.92330028 0.73050595 0.73006854 0.80106838\n",
      " 0.72416795 0.82028392 0.69384798 0.82220737]\n",
      "             exec time: 0:35:51.588351 2021-01-10 01:38:53.788137\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.67979798 0.68650794 0.7031746  0.71251434 0.90503027 0.61666667\n",
      " 0.59136478 0.58065677 0.58204191 0.59835829]\n",
      "             exec time: 0:13:35.732740 2021-01-10 01:52:29.523469\n",
      "---------------------- MinMaxScaler\n",
      "     +++++++ RadiusNeighborsClassifier\n",
      "          scores:  [0.32864358 0.37777778 0.46527778 0.37377451 0.40567766 0.43458625\n",
      " 0.40039683 0.41203704 0.44949495 0.42424242]\n",
      "             exec time: 0:00:32.612522 2021-01-10 01:53:02.139326\n",
      "     +++++++ DecisionTreeClassifier\n",
      "          scores:  [0.73114478 0.73080581 0.73915344 0.66889881 0.82208995 0.75877826\n",
      " 0.77496795 0.69522283 0.79358974 0.70270097]\n",
      "             exec time: 0:00:00.440910 2021-01-10 01:53:02.581156\n",
      "     +++++++ KNeighborsClassifier\n",
      "          scores:  [0.61927825 0.77255639 0.74251216 0.72949735 0.78018926 0.7546841\n",
      " 0.70627706 0.75961779 0.7943219  0.75320513]\n",
      "             exec time: 0:07:48.408273 2021-01-10 02:00:50.991292\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.73395281 0.84483285 0.69459299 0.7512493  0.87007576 0.78182483\n",
      " 0.78808923 0.75833333 0.76076389 0.8337809 ]\n",
      "             exec time: 0:35:18.430679 2021-01-10 02:36:09.423643\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.73842593 0.85403322 0.85313578 0.78541181 0.8451952  0.83891842\n",
      " 0.84613042 0.86056999 0.758061   0.88986102]\n",
      "             exec time: 0:23:42.313556 2021-01-10 02:59:51.739015\n",
      "---------------------- PowerTransformer-yeo-johnson\n",
      "     +++++++ RadiusNeighborsClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:00:12.232360 2021-01-10 03:00:03.972943\n",
      "     +++++++ DecisionTreeClassifier\n",
      "          scores:  [0.72758353 0.71701871 0.63543771 0.7011544  0.76322751 0.72857143\n",
      " 0.78438228 0.64257004 0.7256279  0.66351943]\n",
      "             exec time: 0:00:02.046932 2021-01-10 03:00:06.021621\n",
      "     +++++++ KNeighborsClassifier\n",
      "          scores:  [0.78353175 0.72979798 0.7968254  0.82441829 0.73854016 0.87960454\n",
      " 0.80374255 0.81778147 0.821662   0.6481352 ]\n",
      "             exec time: 0:17:35.616205 2021-01-10 03:17:41.638674\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.73611111 0.76054293 0.81796086 0.76531385 0.72084224 0.8147096\n",
      " 0.76244589 0.77355426 0.86482558 0.73370682]\n",
      "             exec time: 0:36:48.972478 2021-01-10 03:54:30.612676\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.8327381  0.91310541 0.69800385 0.77278185 0.88528139 0.78471881\n",
      " 0.68002137 0.75871212 0.81666667 0.76944444]\n",
      "             exec time: 0:19:02.963119 2021-01-10 04:13:33.577602\n",
      "---------------------- StandardScaler\n",
      "     +++++++ RadiusNeighborsClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:00:12.066308 2021-01-10 04:13:45.644848\n",
      "     +++++++ DecisionTreeClassifier\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "          scores:  [0.78806175 0.77014652 0.74592986 0.75414397 0.67731805 0.77564387\n",
      " 0.72997835 0.72890443 0.68584949 0.75460423]\n",
      "             exec time: 0:00:01.687057 2021-01-10 04:13:47.333877\n",
      "     +++++++ KNeighborsClassifier\n",
      "          scores:  [0.74886364 0.76712793 0.82717949 0.66082112 0.73201459 0.79857143\n",
      " 0.79179293 0.72336601 0.72079772 0.72521368]\n",
      "             exec time: 0:07:55.355176 2021-01-10 04:21:42.690712\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.81111111 0.82587413 0.78392857 0.77063492 0.77146465 0.84226884\n",
      " 0.8281746  0.77436869 0.71085859 0.72444756]\n",
      "             exec time: 0:35:07.307638 2021-01-10 04:56:50.000470\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.76215278 0.86809537 0.73757392 0.777402   0.81649832 0.87772616\n",
      " 0.92795699 0.8625     0.78048341 0.81867409]\n",
      "             exec time: 0:19:34.839767 2021-01-10 05:16:24.842354\n",
      "---------------------- MaxAbsScaler\n",
      "     +++++++ RadiusNeighborsClassifier\n",
      "          scores:  [0.32222222 0.36837607 0.42037037 0.46666667 0.39880952 0.32326007\n",
      " 0.43055556 0.42989418 0.50189394 0.46705147]\n",
      "             exec time: 0:00:27.941126 2021-01-10 05:16:52.784436\n",
      "     +++++++ DecisionTreeClassifier\n",
      "          scores:  [0.6954683  0.80815851 0.70962033 0.70572391 0.61004942 0.76323721\n",
      " 0.84319014 0.79377104 0.77821351 0.71572105]\n",
      "             exec time: 0:00:00.388311 2021-01-10 05:16:53.174350\n",
      "     +++++++ KNeighborsClassifier\n",
      "          scores:  [0.78554361 0.85164141 0.7121538  0.74371596 0.64922112 0.74155825\n",
      " 0.70930344 0.77400285 0.79575163 0.70952381]\n",
      "             exec time: 0:08:00.812180 2021-01-10 05:24:53.988165\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.84555322 0.81899802 0.92165242 0.89705882 0.79140295 0.82326389\n",
      " 0.75108919 0.67222222 0.81048581 0.71102187]\n",
      "             exec time: 0:35:31.186012 2021-01-10 06:00:25.175066\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.81414141 0.81302609 0.69425287 0.83623321 0.87207792 0.87562563\n",
      " 0.81045752 0.622151   0.75176653 0.79545455]\n",
      "             exec time: 0:27:48.213446 2021-01-10 06:28:13.390185\n",
      "---------------------- Normalizer\n",
      "     +++++++ RadiusNeighborsClassifier\n",
      "          scores:  [0.40501655 0.44986264 0.5        0.44444444 0.42424242 0.36820124\n",
      " 0.5125     0.4375     0.46801347 0.38095238]\n",
      "             exec time: 0:00:28.177945 2021-01-10 06:28:41.569615\n",
      "     +++++++ DecisionTreeClassifier\n",
      "          scores:  [0.71425635 0.75142301 0.80918156 0.79981685 0.66128618 0.76034483\n",
      " 0.53861208 0.70606061 0.7391258  0.784375  ]\n",
      "             exec time: 0:00:00.379526 2021-01-10 06:28:41.950120\n",
      "     +++++++ KNeighborsClassifier\n",
      "          scores:  [0.82514881 0.6730464  0.77759972 0.76232957 0.77287157 0.69703952\n",
      " 0.71388889 0.76910774 0.80728231 0.84316239]\n",
      "             exec time: 0:07:51.990015 2021-01-10 06:36:33.941848\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.7490676  0.78153236 0.66881359 0.72852757 0.89110276 0.75645466\n",
      " 0.71645401 0.75107272 0.80030525 0.86706349]\n",
      "             exec time: 0:35:39.993934 2021-01-10 07:12:13.937550\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.74845969 0.73967429 0.8392199  0.68840049 0.84924242 0.73463481\n",
      " 0.8023088  0.78989899 0.66287879 0.79128883]\n",
      "             exec time: 0:23:19.149344 2021-01-10 07:35:33.087797\n",
      "---------------------- QuantileTransformer-normal\n",
      "     +++++++ RadiusNeighborsClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:00:12.243560 2021-01-10 07:35:45.332180\n",
      "     +++++++ DecisionTreeClassifier\n",
      "          scores:  [0.70252506 0.72100122 0.80598545 0.72124183 0.73541667 0.83080808\n",
      " 0.7129329  0.84735265 0.85183535 0.85006105]\n",
      "             exec time: 0:00:02.034727 2021-01-10 07:35:47.368675\n",
      "     +++++++ KNeighborsClassifier\n",
      "          scores:  [0.83127706 0.76965796 0.76752137 0.71757519 0.74608358 0.84296797\n",
      " 0.80138889 0.8531031  0.77205882 0.82401961]\n",
      "             exec time: 0:15:20.096253 2021-01-10 07:51:07.467071\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.78095238 0.82844769 0.83624709 0.73107036 0.79768711 0.70230079\n",
      " 0.72945445 0.77773455 0.81734007 0.8248366 ]\n",
      "             exec time: 0:36:20.066793 2021-01-10 08:27:27.536062\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.77356276 0.7685008  0.80008418 0.69149832 0.79826312 0.80324074\n",
      " 0.83769063 0.81536427 0.69871032 0.7047619 ]\n",
      "             exec time: 0:21:43.455818 2021-01-10 08:49:10.992747\n"
     ]
    }
   ],
   "source": [
    "# let's perform cross validation\n",
    "dfResults = cvAll(X, y, pipelinesMatrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-10T07:49:11.009076Z",
     "start_time": "2021-01-10T07:49:11.000979Z"
    }
   },
   "outputs": [],
   "source": [
    "# summary of ML alorithms screening, let's save it to the file\n",
    "\n",
    "dfResults.to_csv(\"2-results/all-20210103-evaluation_.csv\", sep=\";\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-10T07:49:11.037695Z",
     "start_time": "2021-01-10T07:49:11.012140Z"
    },
    "collapsed": true
   },
   "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>scaler</th>\n",
       "      <th>classifier</th>\n",
       "      <th>balanced_accuracy</th>\n",
       "      <th>balanced_accuracy STD</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NoScaling</td>\n",
       "      <td>RadiusNeighborsClassifier</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NoScaling</td>\n",
       "      <td>DecisionTreeClassifier</td>\n",
       "      <td>0.735504</td>\n",
       "      <td>0.066809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NoScaling</td>\n",
       "      <td>KNeighborsClassifier</td>\n",
       "      <td>0.754814</td>\n",
       "      <td>0.031418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NoScaling</td>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>0.792489</td>\n",
       "      <td>0.053775</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NoScaling</td>\n",
       "      <td>MLPClassifier</td>\n",
       "      <td>0.734573</td>\n",
       "      <td>0.200198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>KBinsDiscretizer-ordinal</td>\n",
       "      <td>RadiusNeighborsClassifier</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>KBinsDiscretizer-ordinal</td>\n",
       "      <td>DecisionTreeClassifier</td>\n",
       "      <td>0.682355</td>\n",
       "      <td>0.060403</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>KBinsDiscretizer-ordinal</td>\n",
       "      <td>KNeighborsClassifier</td>\n",
       "      <td>0.717454</td>\n",
       "      <td>0.049820</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>KBinsDiscretizer-ordinal</td>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>0.723898</td>\n",
       "      <td>0.053936</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>KBinsDiscretizer-ordinal</td>\n",
       "      <td>MLPClassifier</td>\n",
       "      <td>0.750053</td>\n",
       "      <td>0.049435</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>KBinsDiscretizer-onehot</td>\n",
       "      <td>RadiusNeighborsClassifier</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>KBinsDiscretizer-onehot</td>\n",
       "      <td>DecisionTreeClassifier</td>\n",
       "      <td>0.685147</td>\n",
       "      <td>0.064854</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>KBinsDiscretizer-onehot</td>\n",
       "      <td>KNeighborsClassifier</td>\n",
       "      <td>0.656481</td>\n",
       "      <td>0.057560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>KBinsDiscretizer-onehot</td>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>0.674387</td>\n",
       "      <td>0.055339</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>KBinsDiscretizer-onehot</td>\n",
       "      <td>MLPClassifier</td>\n",
       "      <td>0.737426</td>\n",
       "      <td>0.046304</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PolynomialFeatures-2</td>\n",
       "      <td>RadiusNeighborsClassifier</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PolynomialFeatures-2</td>\n",
       "      <td>DecisionTreeClassifier</td>\n",
       "      <td>0.695461</td>\n",
       "      <td>0.066156</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PolynomialFeatures-2</td>\n",
       "      <td>KNeighborsClassifier</td>\n",
       "      <td>0.778168</td>\n",
       "      <td>0.063106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PolynomialFeatures-2</td>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>0.751413</td>\n",
       "      <td>0.067879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PolynomialFeatures-2</td>\n",
       "      <td>MLPClassifier</td>\n",
       "      <td>0.702221</td>\n",
       "      <td>0.066297</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PolynomialFeatures-interaction_only</td>\n",
       "      <td>RadiusNeighborsClassifier</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PolynomialFeatures-interaction_only</td>\n",
       "      <td>DecisionTreeClassifier</td>\n",
       "      <td>0.718284</td>\n",
       "      <td>0.059106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PolynomialFeatures-interaction_only</td>\n",
       "      <td>KNeighborsClassifier</td>\n",
       "      <td>0.761153</td>\n",
       "      <td>0.043303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PolynomialFeatures-interaction_only</td>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>0.777672</td>\n",
       "      <td>0.065556</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PolynomialFeatures-interaction_only</td>\n",
       "      <td>MLPClassifier</td>\n",
       "      <td>0.665611</td>\n",
       "      <td>0.093916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MinMaxScaler</td>\n",
       "      <td>RadiusNeighborsClassifier</td>\n",
       "      <td>0.407191</td>\n",
       "      <td>0.037985</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MinMaxScaler</td>\n",
       "      <td>DecisionTreeClassifier</td>\n",
       "      <td>0.741735</td>\n",
       "      <td>0.044465</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MinMaxScaler</td>\n",
       "      <td>KNeighborsClassifier</td>\n",
       "      <td>0.741214</td>\n",
       "      <td>0.047131</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MinMaxScaler</td>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>0.781750</td>\n",
       "      <td>0.051311</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MinMaxScaler</td>\n",
       "      <td>MLPClassifier</td>\n",
       "      <td>0.826974</td>\n",
       "      <td>0.046539</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PowerTransformer-yeo-johnson</td>\n",
       "      <td>RadiusNeighborsClassifier</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PowerTransformer-yeo-johnson</td>\n",
       "      <td>DecisionTreeClassifier</td>\n",
       "      <td>0.708909</td>\n",
       "      <td>0.046526</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PowerTransformer-yeo-johnson</td>\n",
       "      <td>KNeighborsClassifier</td>\n",
       "      <td>0.784404</td>\n",
       "      <td>0.061115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PowerTransformer-yeo-johnson</td>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>0.775001</td>\n",
       "      <td>0.042549</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PowerTransformer-yeo-johnson</td>\n",
       "      <td>MLPClassifier</td>\n",
       "      <td>0.791147</td>\n",
       "      <td>0.070030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>StandardScaler</td>\n",
       "      <td>RadiusNeighborsClassifier</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>StandardScaler</td>\n",
       "      <td>DecisionTreeClassifier</td>\n",
       "      <td>0.741058</td>\n",
       "      <td>0.034657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>StandardScaler</td>\n",
       "      <td>KNeighborsClassifier</td>\n",
       "      <td>0.749575</td>\n",
       "      <td>0.045616</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>StandardScaler</td>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>0.784313</td>\n",
       "      <td>0.041423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>StandardScaler</td>\n",
       "      <td>MLPClassifier</td>\n",
       "      <td>0.822906</td>\n",
       "      <td>0.057055</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MaxAbsScaler</td>\n",
       "      <td>RadiusNeighborsClassifier</td>\n",
       "      <td>0.412910</td>\n",
       "      <td>0.057305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MaxAbsScaler</td>\n",
       "      <td>DecisionTreeClassifier</td>\n",
       "      <td>0.742315</td>\n",
       "      <td>0.064567</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MaxAbsScaler</td>\n",
       "      <td>KNeighborsClassifier</td>\n",
       "      <td>0.747242</td>\n",
       "      <td>0.054020</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MaxAbsScaler</td>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>0.804275</td>\n",
       "      <td>0.073372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MaxAbsScaler</td>\n",
       "      <td>MLPClassifier</td>\n",
       "      <td>0.788519</td>\n",
       "      <td>0.075198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Normalizer</td>\n",
       "      <td>RadiusNeighborsClassifier</td>\n",
       "      <td>0.439073</td>\n",
       "      <td>0.044575</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Normalizer</td>\n",
       "      <td>DecisionTreeClassifier</td>\n",
       "      <td>0.726448</td>\n",
       "      <td>0.075901</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Normalizer</td>\n",
       "      <td>KNeighborsClassifier</td>\n",
       "      <td>0.764148</td>\n",
       "      <td>0.052382</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Normalizer</td>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>0.771039</td>\n",
       "      <td>0.063979</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Normalizer</td>\n",
       "      <td>MLPClassifier</td>\n",
       "      <td>0.764601</td>\n",
       "      <td>0.057827</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>QuantileTransformer-normal</td>\n",
       "      <td>RadiusNeighborsClassifier</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>QuantileTransformer-normal</td>\n",
       "      <td>DecisionTreeClassifier</td>\n",
       "      <td>0.777916</td>\n",
       "      <td>0.061025</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>QuantileTransformer-normal</td>\n",
       "      <td>KNeighborsClassifier</td>\n",
       "      <td>0.792565</td>\n",
       "      <td>0.042577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>QuantileTransformer-normal</td>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>0.782607</td>\n",
       "      <td>0.044892</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>QuantileTransformer-normal</td>\n",
       "      <td>MLPClassifier</td>\n",
       "      <td>0.769168</td>\n",
       "      <td>0.049957</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                scaler                 classifier  \\\n",
       "0                            NoScaling  RadiusNeighborsClassifier   \n",
       "0                            NoScaling     DecisionTreeClassifier   \n",
       "0                            NoScaling       KNeighborsClassifier   \n",
       "0                            NoScaling     RandomForestClassifier   \n",
       "0                            NoScaling              MLPClassifier   \n",
       "0             KBinsDiscretizer-ordinal  RadiusNeighborsClassifier   \n",
       "0             KBinsDiscretizer-ordinal     DecisionTreeClassifier   \n",
       "0             KBinsDiscretizer-ordinal       KNeighborsClassifier   \n",
       "0             KBinsDiscretizer-ordinal     RandomForestClassifier   \n",
       "0             KBinsDiscretizer-ordinal              MLPClassifier   \n",
       "0              KBinsDiscretizer-onehot  RadiusNeighborsClassifier   \n",
       "0              KBinsDiscretizer-onehot     DecisionTreeClassifier   \n",
       "0              KBinsDiscretizer-onehot       KNeighborsClassifier   \n",
       "0              KBinsDiscretizer-onehot     RandomForestClassifier   \n",
       "0              KBinsDiscretizer-onehot              MLPClassifier   \n",
       "0                 PolynomialFeatures-2  RadiusNeighborsClassifier   \n",
       "0                 PolynomialFeatures-2     DecisionTreeClassifier   \n",
       "0                 PolynomialFeatures-2       KNeighborsClassifier   \n",
       "0                 PolynomialFeatures-2     RandomForestClassifier   \n",
       "0                 PolynomialFeatures-2              MLPClassifier   \n",
       "0  PolynomialFeatures-interaction_only  RadiusNeighborsClassifier   \n",
       "0  PolynomialFeatures-interaction_only     DecisionTreeClassifier   \n",
       "0  PolynomialFeatures-interaction_only       KNeighborsClassifier   \n",
       "0  PolynomialFeatures-interaction_only     RandomForestClassifier   \n",
       "0  PolynomialFeatures-interaction_only              MLPClassifier   \n",
       "0                         MinMaxScaler  RadiusNeighborsClassifier   \n",
       "0                         MinMaxScaler     DecisionTreeClassifier   \n",
       "0                         MinMaxScaler       KNeighborsClassifier   \n",
       "0                         MinMaxScaler     RandomForestClassifier   \n",
       "0                         MinMaxScaler              MLPClassifier   \n",
       "0         PowerTransformer-yeo-johnson  RadiusNeighborsClassifier   \n",
       "0         PowerTransformer-yeo-johnson     DecisionTreeClassifier   \n",
       "0         PowerTransformer-yeo-johnson       KNeighborsClassifier   \n",
       "0         PowerTransformer-yeo-johnson     RandomForestClassifier   \n",
       "0         PowerTransformer-yeo-johnson              MLPClassifier   \n",
       "0                       StandardScaler  RadiusNeighborsClassifier   \n",
       "0                       StandardScaler     DecisionTreeClassifier   \n",
       "0                       StandardScaler       KNeighborsClassifier   \n",
       "0                       StandardScaler     RandomForestClassifier   \n",
       "0                       StandardScaler              MLPClassifier   \n",
       "0                         MaxAbsScaler  RadiusNeighborsClassifier   \n",
       "0                         MaxAbsScaler     DecisionTreeClassifier   \n",
       "0                         MaxAbsScaler       KNeighborsClassifier   \n",
       "0                         MaxAbsScaler     RandomForestClassifier   \n",
       "0                         MaxAbsScaler              MLPClassifier   \n",
       "0                           Normalizer  RadiusNeighborsClassifier   \n",
       "0                           Normalizer     DecisionTreeClassifier   \n",
       "0                           Normalizer       KNeighborsClassifier   \n",
       "0                           Normalizer     RandomForestClassifier   \n",
       "0                           Normalizer              MLPClassifier   \n",
       "0           QuantileTransformer-normal  RadiusNeighborsClassifier   \n",
       "0           QuantileTransformer-normal     DecisionTreeClassifier   \n",
       "0           QuantileTransformer-normal       KNeighborsClassifier   \n",
       "0           QuantileTransformer-normal     RandomForestClassifier   \n",
       "0           QuantileTransformer-normal              MLPClassifier   \n",
       "\n",
       "   balanced_accuracy  balanced_accuracy STD  \n",
       "0                NaN                    NaN  \n",
       "0           0.735504               0.066809  \n",
       "0           0.754814               0.031418  \n",
       "0           0.792489               0.053775  \n",
       "0           0.734573               0.200198  \n",
       "0                NaN                    NaN  \n",
       "0           0.682355               0.060403  \n",
       "0           0.717454               0.049820  \n",
       "0           0.723898               0.053936  \n",
       "0           0.750053               0.049435  \n",
       "0                NaN                    NaN  \n",
       "0           0.685147               0.064854  \n",
       "0           0.656481               0.057560  \n",
       "0           0.674387               0.055339  \n",
       "0           0.737426               0.046304  \n",
       "0                NaN                    NaN  \n",
       "0           0.695461               0.066156  \n",
       "0           0.778168               0.063106  \n",
       "0           0.751413               0.067879  \n",
       "0           0.702221               0.066297  \n",
       "0                NaN                    NaN  \n",
       "0           0.718284               0.059106  \n",
       "0           0.761153               0.043303  \n",
       "0           0.777672               0.065556  \n",
       "0           0.665611               0.093916  \n",
       "0           0.407191               0.037985  \n",
       "0           0.741735               0.044465  \n",
       "0           0.741214               0.047131  \n",
       "0           0.781750               0.051311  \n",
       "0           0.826974               0.046539  \n",
       "0                NaN                    NaN  \n",
       "0           0.708909               0.046526  \n",
       "0           0.784404               0.061115  \n",
       "0           0.775001               0.042549  \n",
       "0           0.791147               0.070030  \n",
       "0                NaN                    NaN  \n",
       "0           0.741058               0.034657  \n",
       "0           0.749575               0.045616  \n",
       "0           0.784313               0.041423  \n",
       "0           0.822906               0.057055  \n",
       "0           0.412910               0.057305  \n",
       "0           0.742315               0.064567  \n",
       "0           0.747242               0.054020  \n",
       "0           0.804275               0.073372  \n",
       "0           0.788519               0.075198  \n",
       "0           0.439073               0.044575  \n",
       "0           0.726448               0.075901  \n",
       "0           0.764148               0.052382  \n",
       "0           0.771039               0.063979  \n",
       "0           0.764601               0.057827  \n",
       "0                NaN                    NaN  \n",
       "0           0.777916               0.061025  \n",
       "0           0.792565               0.042577  \n",
       "0           0.782607               0.044892  \n",
       "0           0.769168               0.049957  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfResults"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-10T07:49:12.161436Z",
     "start_time": "2021-01-10T07:49:11.039926Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Image is worth 1024 words. So let's plot!\n",
    "\n",
    "DataConcatGrpupCVPivot = pd.pivot_table(dfResults, index='scaler', columns='classifier', values='balanced_accuracy',\n",
    "                                  margins=True, margins_name='AVG')\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(8,8)) \n",
    "ax = sns.heatmap(DataConcatGrpupCVPivot, cmap=\"YlGnBu\", annot=True, fmt=\".2f\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Let's probe preprocessing methods with MLP"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-10T14:21:04.730053Z",
     "start_time": "2021-01-10T14:21:04.712880Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+KBinsDiscretizer-ordinal\n",
      " MinMaxScaler\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      " StandardScaler\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      "+KBinsDiscretizer-onehot\n",
      " MinMaxScaler\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      " StandardScaler\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      "+PolynomialFeatures-2\n",
      " MinMaxScaler\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      " StandardScaler\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      "+PolynomialFeatures-interaction_only\n",
      " MinMaxScaler\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      " StandardScaler\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      "+PowerTransformer box-cox\n",
      " MinMaxScaler\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n",
      " StandardScaler\n",
      "             RandomForestClassifier\n",
      "             MLPClassifier\n"
     ]
    }
   ],
   "source": [
    "### fine tuned pipeline with only MLP\n",
    "### Scalers\n",
    "\n",
    "dummyScaller = FunctionTransformer()\n",
    "\n",
    "preprops = [\n",
    "    (KBinsDiscretizer(encode='ordinal', n_bins=6), 'KBinsDiscretizer-ordinal'),\n",
    "    (KBinsDiscretizer(encode='onehot', n_bins=6), 'KBinsDiscretizer-onehot'),\n",
    "    (PolynomialFeatures(2),'PolynomialFeatures-2'),\n",
    "    (PolynomialFeatures(interaction_only=True),'PolynomialFeatures-interaction_only'),\n",
    "    (PowerTransformer(method='box-cox'), \"PowerTransformer box-cox\"),\n",
    "]\n",
    "\n",
    "\n",
    "scalers = [\n",
    "    (MinMaxScaler(), \"MinMaxScaler\"),\n",
    "    (StandardScaler(), \"StandardScaler\"),\n",
    "]\n",
    "\n",
    "### classifiers\n",
    "\n",
    "classifiers = [\n",
    "    RandomForestClassifier(),\n",
    "    MLPClassifier(max_iter=2048,),\n",
    "]\n",
    "\n",
    "\n",
    "### classifiers search space\n",
    "\n",
    "\n",
    "searchSpaceDict = {}\n",
    "\n",
    "searchSpaceDict[\"RandomForestClassifier\"] = {'randomforestclassifier__n_estimators': [20,50, 100, 120,150, 175, 200, 250, 300],\n",
    "                                            'randomforestclassifier__max_depth': [None, 5,10,20,30,40,50,80,100,120,200,220,250,300],\n",
    "                                           'randomforestclassifier__max_features': [\"auto\", \"sqrt\", \"log2\"],}\n",
    "\n",
    "searchSpaceDict[\"MLPClassifier\"] = {\n",
    "                            'mlpclassifier__hidden_layer_sizes': [(20,20), (50,50), (50,), (100,),(150,)],\n",
    "                            'mlpclassifier__activation': ['identity', 'logistic', 'tanh', 'relu'],\n",
    "                            'mlpclassifier__solver': ['lbfgs', 'sgd', 'adam'], }\n",
    "\n",
    "## make the pipelines\n",
    "\n",
    "pipelinesMatrix = {}\n",
    "\n",
    "for preprop, prepropDesc in preprops:\n",
    "    print(\"+\" + prepropDesc)\n",
    "    for scaler, scalerDesc in scalers:\n",
    "        prepropFullName = prepropDesc +  \"+\" + scalerDesc\n",
    "        pipelinesMatrix[prepropFullName] = {}\n",
    "        print(\" \" + scalerDesc) \n",
    "        for classifier in classifiers:\n",
    "            print(\"            \", classifier.__class__.__name__)\n",
    "            pipelinesMatrix[prepropFullName][classifier.__class__.__name__] = make_pipeline(preprop, scaler, classifier)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-10T20:30:58.098056Z",
     "start_time": "2021-01-10T14:21:55.193201Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---------------------- KBinsDiscretizer-ordinal+MinMaxScaler\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.72155377 0.65864876 0.67777778 0.78702166 0.87242697 0.70176282\n",
      " 0.7045177  0.8469697  0.65669516 0.61359428]\n",
      "             exec time: 0:36:38.163513 2021-01-10 15:58:33.358212\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.70940007 0.70054834 0.6453596  0.84583583 0.75108543 0.87532468\n",
      " 0.16666667 0.7179181  0.7498366  0.63931624]\n",
      "             exec time: 0:24:49.695816 2021-01-10 16:23:23.054634\n",
      "---------------------- KBinsDiscretizer-ordinal+StandardScaler\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.77306822 0.64840067 0.76954157 0.73097944 0.77394958 0.78808923\n",
      " 0.73280124 0.7616144  0.82162005 0.62349537]\n",
      "             exec time: 0:35:56.961288 2021-01-10 16:59:20.018276\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.62407407 0.61263736 0.71461004 0.78998779 0.81810501 0.73207071\n",
      " 0.78862434 0.64980507 0.76746032 0.7718254 ]\n",
      "             exec time: 0:19:58.283259 2021-01-10 17:19:18.303868\n",
      "---------------------- KBinsDiscretizer-onehot+MinMaxScaler\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:01:12.366047 2021-01-10 17:20:30.670833\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:00:12.413188 2021-01-10 17:20:43.086256\n",
      "---------------------- KBinsDiscretizer-onehot+StandardScaler\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:01:13.033676 2021-01-10 17:21:56.122394\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:00:12.575844 2021-01-10 17:22:08.699984\n",
      "---------------------- PolynomialFeatures-2+MinMaxScaler\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.76755853 0.84513575 0.84410309 0.82108262 0.88400488 0.85370185\n",
      " 0.79419925 0.83731432 0.81022499 0.79312126]\n",
      "             exec time: 0:36:29.185040 2021-01-10 17:58:37.887710\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.79929194 0.76093535 0.75576599 0.81346154 0.74341085 0.78992165\n",
      " 0.72738414 0.81351852 0.76831502 0.87726098]\n",
      "             exec time: 0:28:38.849484 2021-01-10 18:27:16.738205\n",
      "---------------------- PolynomialFeatures-2+StandardScaler\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.73805361 0.76789006 0.79523327 0.91855159 0.85261438 0.76709803\n",
      " 0.82969577 0.83358135 0.84054834 0.79953704]\n",
      "             exec time: 0:36:33.812178 2021-01-10 19:03:50.553646\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.83696489 0.71920024 0.80573477 0.84678363 0.81151811 0.73297405\n",
      " 0.74686293 0.84842504 0.74224387 0.81146904]\n",
      "             exec time: 0:21:51.051302 2021-01-10 19:25:41.606189\n",
      "---------------------- PolynomialFeatures-interaction_only+MinMaxScaler\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.76230159 0.78741052 0.79329004 0.7368355  0.80363949 0.67106227\n",
      " 0.78111812 0.8022815  0.70472583 0.89844877]\n",
      "             exec time: 0:36:29.873762 2021-01-10 20:02:11.482277\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.6849019  0.77057198 0.85465135 0.88734917 0.74826389 0.8764156\n",
      " 0.75484254 0.89155474 0.64835859 0.82916667]\n",
      "             exec time: 0:26:48.260271 2021-01-10 20:28:59.744557\n",
      "---------------------- PolynomialFeatures-interaction_only+StandardScaler\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [0.78884712 0.81124814 0.8531746  0.73720837 0.77962125 0.70253194\n",
      " 0.7009324  0.67987273 0.76323721 0.68530702]\n",
      "             exec time: 0:37:34.380045 2021-01-10 21:06:34.126540\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [0.84371945 0.7656746  0.8076656  0.72029915 0.82470539 0.89477124\n",
      " 0.76199495 0.85407719 0.79678363 0.66274929]\n",
      "             exec time: 0:22:59.388956 2021-01-10 21:29:33.517755\n",
      "---------------------- PowerTransformer box-cox+MinMaxScaler\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:00:35.384902 2021-01-10 21:30:08.904054\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:00:06.835221 2021-01-10 21:30:15.741560\n",
      "---------------------- PowerTransformer box-cox+StandardScaler\n",
      "     +++++++ RandomForestClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:00:35.464334 2021-01-10 21:30:51.208295\n",
      "     +++++++ MLPClassifier\n",
      "          scores:  [nan nan nan nan nan nan nan nan nan nan]\n",
      "             exec time: 0:00:06.883828 2021-01-10 21:30:58.094241\n"
     ]
    }
   ],
   "source": [
    "# let's perform cross validation\n",
    "dfResults = cvAll(X, y, pipelinesMatrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-10T20:30:58.106986Z",
     "start_time": "2021-01-10T20:30:58.100744Z"
    }
   },
   "outputs": [],
   "source": [
    "# summary of ML alorithms screening, let's save it to the file\n",
    "\n",
    "dfResults.to_csv(\"2-results/all-20210103-evaluation_MLP.csv\", sep=\";\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-10T20:30:58.776952Z",
     "start_time": "2021-01-10T20:30:58.109844Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Image is worth 1024 words. So let's plot!\n",
    "\n",
    "DataConcatGrpupCVPivot = pd.pivot_table(dfResults, index='scaler', columns='classifier', values='balanced_accuracy',\n",
    "                                  margins=True, margins_name='AVG')\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(8,8)) \n",
    "ax = sns.heatmap(DataConcatGrpupCVPivot, cmap=\"YlGnBu\", annot=True, fmt=\".2f\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Conclusions\n",
    "\n",
    "The best parameters so far are:\n",
    "- MLP + MinMaxScaler\n",
    "- +PolynomialFeatures-2 + MinMaxScaler + RandomForestClassifier\n",
    "\n",
    "The first is much more simple, so we will use that one."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-07T07:35:54.631379Z",
     "start_time": "2021-01-07T07:35:54.627341Z"
    }
   },
   "source": [
    "# Let's build the models with the best ML methods\n",
    "\n",
    "- mlp with min-max scaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T20:43:21.150509Z",
     "start_time": "2021-01-18T20:43:21.142534Z"
    }
   },
   "outputs": [],
   "source": [
    "# Here we are splitting input data to train and test set.\n",
    "# default test_size = 0.25\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, stratify=y, shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T20:43:21.158974Z",
     "start_time": "2021-01-18T20:43:21.154215Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Size of the training set: 180\n",
      "Size of the testing set: 60\n"
     ]
    }
   ],
   "source": [
    "print(\"Size of the training set:\", len(X_train))\n",
    "print(\"Size of the testing set:\", len(X_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T20:48:46.370184Z",
     "start_time": "2021-01-18T20:43:21.164742Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(estimator=Pipeline(steps=[('minmaxscaler', MinMaxScaler()),\n",
       "                                       ('mlpclassifier',\n",
       "                                        MLPClassifier(max_iter=20000))]),\n",
       "             n_jobs=4,\n",
       "             param_grid={'mlpclassifier__activation': ['identity', 'logistic',\n",
       "                                                       'tanh', 'relu'],\n",
       "                         'mlpclassifier__hidden_layer_sizes': [(20,), (20, 20),\n",
       "                                                               (20, 20, 20),\n",
       "                                                               (30, 20),\n",
       "                                                               (50, 50), (50,),\n",
       "                                                               (100,), (150,)],\n",
       "                         'mlpclassifier__solver': ['lbfgs', 'sgd', 'adam']},\n",
       "             scoring='balanced_accuracy')"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# lets build models with best pipelines\n",
    "\n",
    "searchSpaceDict = {}\n",
    "searchSpaceDict[\"MLPClassifier\"] = {\n",
    "                            'mlpclassifier__hidden_layer_sizes': [(20,), (20,20), (20,20,20), (30,20), (50,50), (50,), (100,),(150,)],\n",
    "                            'mlpclassifier__activation': ['identity', 'logistic', 'tanh', 'relu'],\n",
    "                            'mlpclassifier__solver': ['lbfgs', 'sgd', 'adam'],\n",
    "#                             'mlpclassifier__learning_rate': ['constant', 'invscaling', 'adaptive'],\n",
    "#                             'mlpclassifier__momentum': [0, 0.9],\n",
    "#                             'mlpclassifier__learning_rate_init': [0.01, 0.2],\n",
    "                            }\n",
    "\n",
    "scaler = MinMaxScaler()\n",
    "scalerPoly= PolynomialFeatures(2)\n",
    "classifier = MLPClassifier(max_iter=20000,)\n",
    "searchSpace = searchSpaceDict[\"MLPClassifier\"]\n",
    "\n",
    "# pipeline = make_pipeline(scalerPoly, scaler, classifier)\n",
    "pipeline = make_pipeline(scaler, classifier)\n",
    "\n",
    "clf = GridSearchCV(pipeline, searchSpace, scoring='balanced_accuracy', n_jobs=4)\n",
    "\n",
    "clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T20:48:46.379458Z",
     "start_time": "2021-01-18T20:48:46.373531Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'mlpclassifier__activation': 'tanh',\n",
       " 'mlpclassifier__hidden_layer_sizes': (30, 20),\n",
       " 'mlpclassifier__solver': 'lbfgs'}"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# print best params\n",
    "\n",
    "clf.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T20:48:46.395995Z",
     "start_time": "2021-01-18T20:48:46.383228Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.85"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# tmake predictions for best params, print score\n",
    "\n",
    "y_test_pred = clf.best_estimator_.predict(X_test)\n",
    "\n",
    "clf.best_estimator_.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T20:48:46.418426Z",
     "start_time": "2021-01-18T20:48:46.398618Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification report for classifier GridSearchCV(estimator=Pipeline(steps=[('minmaxscaler', MinMaxScaler()),\n",
      "                                       ('mlpclassifier',\n",
      "                                        MLPClassifier(max_iter=20000))]),\n",
      "             n_jobs=4,\n",
      "             param_grid={'mlpclassifier__activation': ['identity', 'logistic',\n",
      "                                                       'tanh', 'relu'],\n",
      "                         'mlpclassifier__hidden_layer_sizes': [(20,), (20, 20),\n",
      "                                                               (20, 20, 20),\n",
      "                                                               (30, 20),\n",
      "                                                               (50, 50), (50,),\n",
      "                                                               (100,), (150,)],\n",
      "                         'mlpclassifier__solver': ['lbfgs', 'sgd', 'adam']},\n",
      "             scoring='balanced_accuracy'):\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  bathroom_d       1.00      0.67      0.80         3\n",
      "  bathroom_g       0.89      1.00      0.94         8\n",
      " living_room       0.85      0.88      0.86        25\n",
      "      office       0.80      0.89      0.84         9\n",
      "      room_k       1.00      1.00      1.00         8\n",
      "      room_m       0.60      0.43      0.50         7\n",
      "\n",
      "    accuracy                           0.85        60\n",
      "   macro avg       0.86      0.81      0.82        60\n",
      "weighted avg       0.84      0.85      0.84        60\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(f\"Classification report for classifier {clf}:\\n\"\n",
    "      f\"{metrics.classification_report(y_test, y_test_pred)}\\n\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T20:48:46.924485Z",
     "start_time": "2021-01-18T20:48:46.420560Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(10, 10))\n",
    "sns.set(font_scale=2.0) # Adjust to fit\n",
    "cm = metrics.plot_confusion_matrix(clf, X_test, y_test, ax=ax)\n",
    "\n",
    "ax.tick_params(axis='both', which='major', labelsize=10)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-10T20:48:29.606979Z",
     "start_time": "2021-01-10T20:48:29.604467Z"
    }
   },
   "source": [
    "# Save models for future use"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T20:48:46.931041Z",
     "start_time": "2021-01-18T20:48:46.926919Z"
    }
   },
   "outputs": [],
   "source": [
    "import joblib"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T20:48:46.946944Z",
     "start_time": "2021-01-18T20:48:46.934246Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['3-models/20200118-MLP.pkl']"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "filename = '3-models/20200118-MLP.pkl'\n",
    "joblib.dump(clf.best_estimator_, filename)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T20:48:46.968625Z",
     "start_time": "2021-01-18T20:48:46.951339Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.85\n"
     ]
    }
   ],
   "source": [
    "# test of the loaded models\n",
    "\n",
    "# load the model from disk\n",
    "loaded_model = joblib.load(filename)\n",
    "result = loaded_model.score(X_test, y_test)\n",
    "print(result)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Reuse saved models for new predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T20:55:21.383678Z",
     "start_time": "2021-01-18T20:55:21.371097Z"
    }
   },
   "outputs": [],
   "source": [
    "import joblib\n",
    "import numpy as np\n",
    "\n",
    "def localizeTheCat(rssis):\n",
    "    # labels = room ids\n",
    "    labels = loaded_model.classes_.tolist()\n",
    "\n",
    "    # where is the cat - location\n",
    "    where_is_a_cat = loaded_model.predict(rssis)[0].tolist()\n",
    "\n",
    "    # probabilities for each room\n",
    "    where_is_a_cat_p = loaded_model.predict_proba(rssis)[0].tolist()\n",
    "\n",
    "    # prediction probability:\n",
    "    location_p = round(where_is_a_cat_p[labels.index(where_is_a_cat)] * 100, 1)\n",
    "    return (where_is_a_cat, location_p)\n",
    "\n",
    "\n",
    "# load models\n",
    "filename = '3-models/20200118-MLP.pkl'\n",
    "loaded_model = joblib.load(filename)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T20:55:23.582872Z",
     "start_time": "2021-01-18T20:55:23.576199Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The cat is in the office with p=100.0%\n"
     ]
    }
   ],
   "source": [
    "# input RSSI array, fetch from the database or file with last measurements\n",
    "rssis = np.array([[-70, -70, -30, -110]])\n",
    "\n",
    "# let's make a prediction!\n",
    "where_is_a_cat, location_p = localizeTheCat(rssis)\n",
    "\n",
    "# and output the results\n",
    "print(\"The cat is in the %s with p=%s%%\" % (where_is_a_cat, location_p))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-18T20:55:27.124542Z",
     "start_time": "2021-01-18T20:55:27.122183Z"
    }
   },
   "outputs": [],
   "source": [
    "# and this is it! thanks for watching!"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.6"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": true,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}