trainer.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "db45dff1-0262-43b3-b911-f082e402e634",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.linear_model import LogisticRegression, RidgeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier\n",
    "from sklearn.metrics import accuracy_score\n",
    "import pickle\n",
    "\n",
    "import cv2\n",
    "import numpy as np\n",
    "\n",
    "from mediapipe.python.solutions import drawing_utils as mp_drawing\n",
    "from mediapipe.python.solutions import holistic as mp_holistic"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "5de97d19-b746-4203-adaf-970b3aca3b0e",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('coordinates.csv')\n",
    "df.dropna(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "18b7852d-e743-483c-86f9-737856e1c46d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>class</th>\n",
       "      <th>x1</th>\n",
       "      <th>y1</th>\n",
       "      <th>z1</th>\n",
       "      <th>v1</th>\n",
       "      <th>x2</th>\n",
       "      <th>y2</th>\n",
       "      <th>z2</th>\n",
       "      <th>v2</th>\n",
       "      <th>x3</th>\n",
       "      <th>...</th>\n",
       "      <th>z499</th>\n",
       "      <th>v499</th>\n",
       "      <th>x500</th>\n",
       "      <th>y500</th>\n",
       "      <th>z500</th>\n",
       "      <th>v500</th>\n",
       "      <th>x501</th>\n",
       "      <th>y501</th>\n",
       "      <th>z501</th>\n",
       "      <th>v501</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>onphone</td>\n",
       "      <td>0.328635</td>\n",
       "      <td>0.429501</td>\n",
       "      <td>-0.545699</td>\n",
       "      <td>0.999715</td>\n",
       "      <td>0.357870</td>\n",
       "      <td>0.377662</td>\n",
       "      <td>-0.513882</td>\n",
       "      <td>0.998970</td>\n",
       "      <td>0.373069</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.003363</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.378759</td>\n",
       "      <td>0.367419</td>\n",
       "      <td>0.002722</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.383056</td>\n",
       "      <td>0.360641</td>\n",
       "      <td>0.002427</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>onphone</td>\n",
       "      <td>0.328712</td>\n",
       "      <td>0.428614</td>\n",
       "      <td>-0.615895</td>\n",
       "      <td>0.999713</td>\n",
       "      <td>0.357924</td>\n",
       "      <td>0.377663</td>\n",
       "      <td>-0.572709</td>\n",
       "      <td>0.998949</td>\n",
       "      <td>0.373282</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.004115</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.379162</td>\n",
       "      <td>0.364425</td>\n",
       "      <td>0.002475</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.383338</td>\n",
       "      <td>0.359364</td>\n",
       "      <td>0.002037</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>onphone</td>\n",
       "      <td>0.328689</td>\n",
       "      <td>0.427790</td>\n",
       "      <td>-0.632160</td>\n",
       "      <td>0.999718</td>\n",
       "      <td>0.357870</td>\n",
       "      <td>0.377635</td>\n",
       "      <td>-0.587004</td>\n",
       "      <td>0.998951</td>\n",
       "      <td>0.373285</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.004626</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.381348</td>\n",
       "      <td>0.362481</td>\n",
       "      <td>0.001779</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.385641</td>\n",
       "      <td>0.357448</td>\n",
       "      <td>0.001243</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>onphone</td>\n",
       "      <td>0.328687</td>\n",
       "      <td>0.428207</td>\n",
       "      <td>-0.555022</td>\n",
       "      <td>0.999706</td>\n",
       "      <td>0.357885</td>\n",
       "      <td>0.378165</td>\n",
       "      <td>-0.516839</td>\n",
       "      <td>0.998898</td>\n",
       "      <td>0.373417</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.004487</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.382253</td>\n",
       "      <td>0.362935</td>\n",
       "      <td>0.002303</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.386400</td>\n",
       "      <td>0.358029</td>\n",
       "      <td>0.001816</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>onphone</td>\n",
       "      <td>0.328650</td>\n",
       "      <td>0.429459</td>\n",
       "      <td>-0.549788</td>\n",
       "      <td>0.999651</td>\n",
       "      <td>0.357896</td>\n",
       "      <td>0.379126</td>\n",
       "      <td>-0.514216</td>\n",
       "      <td>0.998677</td>\n",
       "      <td>0.373497</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.003797</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.383240</td>\n",
       "      <td>0.363496</td>\n",
       "      <td>0.002815</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.387517</td>\n",
       "      <td>0.358570</td>\n",
       "      <td>0.002309</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 2005 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     class        x1        y1        z1        v1        x2        y2  \\\n",
       "0  onphone  0.328635  0.429501 -0.545699  0.999715  0.357870  0.377662   \n",
       "1  onphone  0.328712  0.428614 -0.615895  0.999713  0.357924  0.377663   \n",
       "2  onphone  0.328689  0.427790 -0.632160  0.999718  0.357870  0.377635   \n",
       "3  onphone  0.328687  0.428207 -0.555022  0.999706  0.357885  0.378165   \n",
       "4  onphone  0.328650  0.429459 -0.549788  0.999651  0.357896  0.379126   \n",
       "\n",
       "         z2        v2        x3  ...      z499  v499      x500      y500  \\\n",
       "0 -0.513882  0.998970  0.373069  ... -0.003363   0.0  0.378759  0.367419   \n",
       "1 -0.572709  0.998949  0.373282  ... -0.004115   0.0  0.379162  0.364425   \n",
       "2 -0.587004  0.998951  0.373285  ... -0.004626   0.0  0.381348  0.362481   \n",
       "3 -0.516839  0.998898  0.373417  ... -0.004487   0.0  0.382253  0.362935   \n",
       "4 -0.514216  0.998677  0.373497  ... -0.003797   0.0  0.383240  0.363496   \n",
       "\n",
       "       z500  v500      x501      y501      z501  v501  \n",
       "0  0.002722   0.0  0.383056  0.360641  0.002427   0.0  \n",
       "1  0.002475   0.0  0.383338  0.359364  0.002037   0.0  \n",
       "2  0.001779   0.0  0.385641  0.357448  0.001243   0.0  \n",
       "3  0.002303   0.0  0.386400  0.358029  0.001816   0.0  \n",
       "4  0.002815   0.0  0.387517  0.358570  0.002309   0.0  \n",
       "\n",
       "[5 rows x 2005 columns]"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "5c8e3ef6-8b94-439b-98d6-ac2747a88688",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df.drop('class', axis=1)\n",
    "y = df['class']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "40c5983c-e2c6-4bfc-8ae9-725017b956a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=1111)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "3cc4b686-164a-48d4-885f-af7b06467340",
   "metadata": {},
   "outputs": [],
   "source": [
    "pipelines = {\n",
    "    'lr': make_pipeline(StandardScaler(), LogisticRegression()),\n",
    "    'rc': make_pipeline(StandardScaler(), RidgeClassifier()),\n",
    "    'rf': make_pipeline(StandardScaler(), RandomForestClassifier()),\n",
    "    'gb': make_pipeline(StandardScaler(), GradientBoostingClassifier())\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "ed34c540-aa48-4e64-ba17-cdd93187ab77",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n"
     ]
    }
   ],
   "source": [
    "fit_models = {}\n",
    "for algo, pipeline in pipelines.items():\n",
    "    model = pipeline.fit(X_train, y_train)\n",
    "    fit_models[algo] = model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "c47a1c2f-2c95-45d0-91dd-9fa209861f6f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "lr 1.0\n",
      "rc 1.0\n",
      "rf 1.0\n",
      "gb 1.0\n"
     ]
    }
   ],
   "source": [
    "for algo, model in fit_models.items():\n",
    "    yhat = model.predict(X_test)\n",
    "    print(algo, accuracy_score(y_test, yhat))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "1186e8e7-6440-4285-984d-551964f295ae",
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('body_language.pkl', 'wb') as f:\n",
    "    pickle.dump(fit_models['rf'], f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "0f2218db-c484-416e-8f47-0947786a319e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['break', 'onphone', 'onphone', 'sleepy', 'break', 'break',\n",
       "       'sleepy', 'sleepy', 'sleepy', 'sleepy', 'onphone', 'onphone',\n",
       "       'onphone', 'break', 'sleepy', 'onphone', 'break', 'sleepy',\n",
       "       'onphone', 'sleepy', 'onphone', 'sleepy', 'onphone', 'sleepy',\n",
       "       'break', 'sleepy', 'sleepy', 'break', 'onphone', 'onphone',\n",
       "       'sleepy', 'break', 'sleepy', 'break', 'sleepy', 'sleepy', 'sleepy',\n",
       "       'sleepy', 'break', 'break', 'break', 'onphone', 'onphone', 'break',\n",
       "       'sleepy', 'onphone', 'sleepy', 'sleepy', 'break', 'sleepy',\n",
       "       'onphone', 'break', 'onphone', 'break', 'break', 'sleepy', 'break',\n",
       "       'sleepy', 'break', 'onphone', 'break', 'onphone', 'sleepy',\n",
       "       'onphone', 'onphone', 'sleepy', 'break', 'sleepy', 'sleepy',\n",
       "       'break', 'onphone', 'onphone', 'break', 'onphone', 'onphone',\n",
       "       'sleepy', 'break', 'sleepy', 'onphone', 'break', 'sleepy',\n",
       "       'onphone', 'break', 'onphone', 'sleepy', 'onphone', 'sleepy',\n",
       "       'break', 'onphone', 'break', 'sleepy', 'break', 'onphone', 'break',\n",
       "       'sleepy', 'onphone', 'break', 'break', 'onphone', 'break',\n",
       "       'sleepy', 'sleepy', 'sleepy', 'break', 'onphone', 'onphone',\n",
       "       'break', 'break', 'onphone', 'onphone', 'break', 'sleepy',\n",
       "       'onphone', 'break', 'break', 'onphone', 'sleepy', 'sleepy',\n",
       "       'onphone', 'break', 'sleepy', 'onphone', 'break', 'break',\n",
       "       'sleepy', 'break', 'onphone', 'sleepy', 'sleepy', 'break',\n",
       "       'sleepy', 'sleepy', 'break', 'break', 'break', 'break', 'break',\n",
       "       'onphone', 'break', 'break', 'break', 'break', 'onphone', 'sleepy',\n",
       "       'onphone', 'onphone', 'onphone', 'sleepy', 'break', 'onphone',\n",
       "       'break', 'break', 'break', 'break', 'onphone', 'break', 'break',\n",
       "       'break', 'onphone', 'onphone', 'sleepy', 'break', 'sleepy',\n",
       "       'sleepy', 'break', 'onphone', 'onphone', 'break', 'break', 'break',\n",
       "       'sleepy', 'break', 'onphone', 'onphone', 'break', 'break',\n",
       "       'onphone', 'break', 'sleepy', 'break', 'break', 'onphone',\n",
       "       'sleepy', 'sleepy', 'sleepy', 'sleepy', 'onphone', 'onphone',\n",
       "       'break', 'onphone', 'break', 'sleepy', 'sleepy', 'sleepy',\n",
       "       'onphone', 'onphone', 'onphone', 'break', 'sleepy', 'break',\n",
       "       'break', 'sleepy', 'onphone', 'onphone', 'break', 'onphone',\n",
       "       'sleepy', 'onphone', 'onphone', 'break', 'sleepy', 'onphone',\n",
       "       'sleepy', 'sleepy', 'break', 'onphone', 'onphone', 'break',\n",
       "       'sleepy', 'break', 'break', 'break', 'break', 'sleepy', 'sleepy',\n",
       "       'onphone', 'sleepy', 'sleepy', 'break', 'onphone', 'sleepy',\n",
       "       'sleepy', 'break', 'break', 'sleepy', 'sleepy', 'sleepy',\n",
       "       'onphone', 'break', 'sleepy', 'break', 'sleepy', 'sleepy', 'break',\n",
       "       'onphone', 'break', 'sleepy', 'onphone', 'onphone', 'onphone',\n",
       "       'onphone', 'sleepy', 'onphone', 'break', 'onphone', 'sleepy',\n",
       "       'onphone', 'break', 'onphone'], dtype=object)"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fit_models['rf'].predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ab0cbee6-9a16-41a6-8a1b-b19fc50d8e44",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}