02b - ML Classifiers_per_Modality_Chest_Device.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "from sklearn import metrics\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis as LDA\n",
    "from sklearn.ensemble import AdaBoostClassifier, RandomForestClassifier\n",
    "from sklearn.model_selection import cross_validate, LeaveOneGroupOut\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "import sklearn\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Subject Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "BASE_PATH = 'segmented_data'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "# subjects_data = [pd.read_feather(os.path.join(BASE_PATH, subject)) for subject in os.listdir(BASE_PATH)]\n",
    "# df = pd.concat(subjects_data)\n",
    "# df.drop(columns=['index'], inplace=True)\n",
    "# df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df.to_csv('segmented_data/all_subjects.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(f'{BASE_PATH}/all_subjects.csv', index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(121813, 79)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>ACC_x_absint</th>\n",
       "      <th>ACC_x_mean</th>\n",
       "      <th>ACC_x_peak</th>\n",
       "      <th>ACC_x_std</th>\n",
       "      <th>ACC_xyz_absint</th>\n",
       "      <th>ACC_xyz_std</th>\n",
       "      <th>ACC_xzy_mean</th>\n",
       "      <th>ACC_y_absint</th>\n",
       "      <th>ACC_y_mean</th>\n",
       "      <th>ACC_y_peak</th>\n",
       "      <th>...</th>\n",
       "      <th>RESP_stretch</th>\n",
       "      <th>RESP_volume</th>\n",
       "      <th>TEMP_drange</th>\n",
       "      <th>TEMP_max</th>\n",
       "      <th>TEMP_mean</th>\n",
       "      <th>TEMP_min</th>\n",
       "      <th>TEMP_slope</th>\n",
       "      <th>TEMP_std</th>\n",
       "      <th>label</th>\n",
       "      <th>subject</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>33018.685611</td>\n",
       "      <td>0.786178</td>\n",
       "      <td>0.8766</td>\n",
       "      <td>0.030389</td>\n",
       "      <td>13233.532886</td>\n",
       "      <td>21266.015893</td>\n",
       "      <td>6736.963395</td>\n",
       "      <td>6256.039785</td>\n",
       "      <td>0.148957</td>\n",
       "      <td>0.2504</td>\n",
       "      <td>...</td>\n",
       "      <td>10.783828</td>\n",
       "      <td>125.194862</td>\n",
       "      <td>1.010598</td>\n",
       "      <td>35.171936</td>\n",
       "      <td>34.988750</td>\n",
       "      <td>34.803101</td>\n",
       "      <td>-0.000002</td>\n",
       "      <td>0.054536</td>\n",
       "      <td>3</td>\n",
       "      <td>S8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>33012.188012</td>\n",
       "      <td>0.786023</td>\n",
       "      <td>0.8766</td>\n",
       "      <td>0.030383</td>\n",
       "      <td>13228.857486</td>\n",
       "      <td>21267.760722</td>\n",
       "      <td>6732.487596</td>\n",
       "      <td>6260.111985</td>\n",
       "      <td>0.149054</td>\n",
       "      <td>0.2504</td>\n",
       "      <td>...</td>\n",
       "      <td>10.783828</td>\n",
       "      <td>125.194862</td>\n",
       "      <td>1.010598</td>\n",
       "      <td>35.171936</td>\n",
       "      <td>34.988785</td>\n",
       "      <td>34.803101</td>\n",
       "      <td>-0.000002</td>\n",
       "      <td>0.054524</td>\n",
       "      <td>3</td>\n",
       "      <td>S8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>33006.168612</td>\n",
       "      <td>0.785880</td>\n",
       "      <td>0.8766</td>\n",
       "      <td>0.030367</td>\n",
       "      <td>13225.533786</td>\n",
       "      <td>21268.840449</td>\n",
       "      <td>6728.895262</td>\n",
       "      <td>6264.232385</td>\n",
       "      <td>0.149152</td>\n",
       "      <td>0.2504</td>\n",
       "      <td>...</td>\n",
       "      <td>10.783828</td>\n",
       "      <td>125.194862</td>\n",
       "      <td>1.010598</td>\n",
       "      <td>35.171936</td>\n",
       "      <td>34.988869</td>\n",
       "      <td>34.803101</td>\n",
       "      <td>-0.000002</td>\n",
       "      <td>0.054485</td>\n",
       "      <td>3</td>\n",
       "      <td>S8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>33005.078112</td>\n",
       "      <td>0.785855</td>\n",
       "      <td>0.8766</td>\n",
       "      <td>0.030379</td>\n",
       "      <td>13228.145587</td>\n",
       "      <td>21269.264878</td>\n",
       "      <td>6729.222863</td>\n",
       "      <td>6268.463386</td>\n",
       "      <td>0.149253</td>\n",
       "      <td>0.2504</td>\n",
       "      <td>...</td>\n",
       "      <td>10.783828</td>\n",
       "      <td>125.194862</td>\n",
       "      <td>1.010598</td>\n",
       "      <td>35.171936</td>\n",
       "      <td>34.988983</td>\n",
       "      <td>34.803101</td>\n",
       "      <td>-0.000002</td>\n",
       "      <td>0.054432</td>\n",
       "      <td>3</td>\n",
       "      <td>S8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>33008.020712</td>\n",
       "      <td>0.785926</td>\n",
       "      <td>0.9390</td>\n",
       "      <td>0.030564</td>\n",
       "      <td>13234.658287</td>\n",
       "      <td>21268.410145</td>\n",
       "      <td>6732.741796</td>\n",
       "      <td>6270.908585</td>\n",
       "      <td>0.149312</td>\n",
       "      <td>0.2504</td>\n",
       "      <td>...</td>\n",
       "      <td>10.783828</td>\n",
       "      <td>119.178477</td>\n",
       "      <td>1.010598</td>\n",
       "      <td>35.171936</td>\n",
       "      <td>34.989079</td>\n",
       "      <td>34.803101</td>\n",
       "      <td>-0.000002</td>\n",
       "      <td>0.054387</td>\n",
       "      <td>3</td>\n",
       "      <td>S8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 79 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   ACC_x_absint  ACC_x_mean  ACC_x_peak  ACC_x_std  ACC_xyz_absint  \\\n",
       "0  33018.685611    0.786178      0.8766   0.030389    13233.532886   \n",
       "1  33012.188012    0.786023      0.8766   0.030383    13228.857486   \n",
       "2  33006.168612    0.785880      0.8766   0.030367    13225.533786   \n",
       "3  33005.078112    0.785855      0.8766   0.030379    13228.145587   \n",
       "4  33008.020712    0.785926      0.9390   0.030564    13234.658287   \n",
       "\n",
       "    ACC_xyz_std  ACC_xzy_mean  ACC_y_absint  ACC_y_mean  ACC_y_peak  ...  \\\n",
       "0  21266.015893   6736.963395   6256.039785    0.148957      0.2504  ...   \n",
       "1  21267.760722   6732.487596   6260.111985    0.149054      0.2504  ...   \n",
       "2  21268.840449   6728.895262   6264.232385    0.149152      0.2504  ...   \n",
       "3  21269.264878   6729.222863   6268.463386    0.149253      0.2504  ...   \n",
       "4  21268.410145   6732.741796   6270.908585    0.149312      0.2504  ...   \n",
       "\n",
       "   RESP_stretch  RESP_volume  TEMP_drange   TEMP_max  TEMP_mean   TEMP_min  \\\n",
       "0     10.783828   125.194862     1.010598  35.171936  34.988750  34.803101   \n",
       "1     10.783828   125.194862     1.010598  35.171936  34.988785  34.803101   \n",
       "2     10.783828   125.194862     1.010598  35.171936  34.988869  34.803101   \n",
       "3     10.783828   125.194862     1.010598  35.171936  34.988983  34.803101   \n",
       "4     10.783828   119.178477     1.010598  35.171936  34.989079  34.803101   \n",
       "\n",
       "   TEMP_slope  TEMP_std  label  subject  \n",
       "0   -0.000002  0.054536      3       S8  \n",
       "1   -0.000002  0.054524      3       S8  \n",
       "2   -0.000002  0.054485      3       S8  \n",
       "3   -0.000002  0.054432      3       S8  \n",
       "4   -0.000002  0.054387      3       S8  \n",
       "\n",
       "[5 rows x 79 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['ACC_x_absint', 'ACC_x_mean', 'ACC_x_peak', 'ACC_x_std',\n",
       "       'ACC_xyz_absint', 'ACC_xyz_std', 'ACC_xzy_mean', 'ACC_y_absint',\n",
       "       'ACC_y_mean', 'ACC_y_peak', 'ACC_y_std', 'ACC_z_absint', 'ACC_z_mean',\n",
       "       'ACC_z_peak', 'ACC_z_std', 'ECG_hr_mean', 'ECG_hr_std', 'ECG_hrv_HF',\n",
       "       'ECG_hrv_HFn', 'ECG_hrv_LF', 'ECG_hrv_LFHF_ratio', 'ECG_hrv_LFn',\n",
       "       'ECG_hrv_NN50', 'ECG_hrv_RMS', 'ECG_hrv_TINN', 'ECG_hrv_ULF',\n",
       "       'ECG_hrv_VHF', 'ECG_hrv_f_sum', 'ECG_hrv_pNN50', 'ECG_hrv_rel_HF',\n",
       "       'ECG_hrv_rel_LF', 'ECG_hrv_rel_ULF', 'ECG_hrv_rel_VHF', 'EDA_SCL_mean',\n",
       "       'EDA_SCL_std', 'EDA_SCR_mean', 'EDA_SCR_no', 'EDA_SCR_std',\n",
       "       'EDA_ampl_sum', 'EDA_corr_SCL_t', 'EDA_drange', 'EDA_max', 'EDA_mean',\n",
       "       'EDA_min', 'EDA_scr_area', 'EDA_slope', 'EDA_std', 'EMG_absint',\n",
       "       'EMG_drange', 'EMG_mean', 'EMG_mean_freq', 'EMG_median',\n",
       "       'EMG_median_freq', 'EMG_peak_amp_mean', 'EMG_peak_amp_norm_sum',\n",
       "       'EMG_peak_amp_std', 'EMG_peak_amp_sum', 'EMG_peak_freq', 'EMG_peak_no',\n",
       "       'EMG_perc_10', 'EMG_perc_90', 'EMG_std', 'RESP_breath_rate',\n",
       "       'RESP_duration', 'RESP_exhal_mean', 'RESP_exhal_std', 'RESP_ie_ratio',\n",
       "       'RESP_inhal_mean', 'RESP_inhal_std', 'RESP_stretch', 'RESP_volume',\n",
       "       'TEMP_drange', 'TEMP_max', 'TEMP_mean', 'TEMP_min', 'TEMP_slope',\n",
       "       'TEMP_std', 'label', 'subject'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Picking a modality\n",
    "\n",
    "Prior executing the models, uncomment a modality from the following cells"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {},
   "outputs": [],
   "source": [
    "# modality = 'ACC'\n",
    "# modality = 'ECG'\n",
    "# modality = 'EDA'\n",
    "# modality = 'EMG'\n",
    "# modality = 'RESP'\n",
    "modality = 'TEMP'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_report = pd.DataFrame()\n",
    "df_report.loc['ACC', :] = []\n",
    "df_report.loc['ECG', :] = []\n",
    "df_report.loc['EDA', :] = []\n",
    "df_report.loc['EMG', :] = []\n",
    "df_report.loc['RESP', :] = []\n",
    "df_report.loc['TEMP', :] = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {},
   "outputs": [],
   "source": [
    "mod_cols = df.columns[df.columns.str.startswith(modality)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Splitting data for CV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(113425,)\n",
      "(113621,)\n",
      "(113628,)\n",
      "(113624,)\n",
      "(113601,)\n",
      "(113648,)\n",
      "(113429,)\n",
      "(114049,)\n",
      "(113913,)\n",
      "(113872,)\n",
      "(113665,)\n",
      "(113725,)\n",
      "(113740,)\n",
      "(113697,)\n",
      "(113745,)\n"
     ]
    }
   ],
   "source": [
    "X = df[mod_cols]\n",
    "y = df['label']\n",
    "y_b = y.copy().apply(lambda x: 1 if x == 2 else 0)  # stress vs non-stress\n",
    "loso_cv = LeaveOneGroupOut()  # just to analyze training set sizes\n",
    "for train, test in loso_cv.split(X, y, groups=df['subject']):\n",
    "    print(train.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(121813, 6)"
      ]
     },
     "execution_count": 194,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Hyperparameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_estimators = 100\n",
    "min_samples_split = 20\n",
    "criterion = 'entropy'  # information gain\n",
    "subjects = df['subject'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {},
   "outputs": [],
   "source": [
    "def run_cv(clf, X, y, groups, cv, scoring=['accuracy', 'f1_macro'], return_train_score=True,\n",
    "          return_estimator=True, n_jobs=-1):\n",
    "    \"\"\"\n",
    "        More on cross validation: https://scikit-learn.org/stable/modules/cross_validation.html#\n",
    "        More on scoring: https://scikit-learn.org/stable/modules/model_evaluation.html#scoring-parameter\n",
    "    \"\"\"\n",
    "    return cross_validate(clf, X, y, groups=groups, cv=cv, scoring=scoring, return_train_score=return_train_score,\n",
    "                          return_estimator=return_estimator, n_jobs=n_jobs)\n",
    "def raw_scores(cv):\n",
    "    return np.mean(cv['test_accuracy']), np.mean(cv['test_f1_macro'])\n",
    "\n",
    "def test_score(cv):\n",
    "    return {\n",
    "        'accuracy': f\"{np.mean(cv['test_accuracy'])} +-{np.std(cv['test_accuracy'])}\",\n",
    "        'f1_macro': f\"{np.mean(cv['test_f1_macro'])} +-{np.std(cv['test_f1_macro'])}\"\n",
    "    }\n",
    "\n",
    "def min_max(acc_scores):\n",
    "    print(np.argmax(acc_scores), np.amax(acc_scores))\n",
    "    print(np.argmin(acc_scores), np.amin(acc_scores))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_rf_pipeline(n_estimators=100, min_samples_split=20, criterion='entropy', n_jobs=-1):\n",
    "    model_rf =  RandomForestClassifier(n_estimators=n_estimators, min_samples_split=min_samples_split, \n",
    "                                  criterion=criterion, n_jobs=-1)\n",
    "    return make_pipeline(StandardScaler(), model_rf)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Baseline vs Stress vs Amusement"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'accuracy': '0.499754647995666 +-0.19306057821544867',\n",
       " 'f1_macro': '0.40883665575794165 +-0.18470415760962317'}"
      ]
     },
     "execution_count": 198,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_rf_all = create_rf_pipeline(n_estimators=n_estimators, min_samples_split=min_samples_split,\n",
    "                                criterion=criterion, n_jobs=-1)\n",
    "cv_rf_all = run_cv(clf_rf_all, X, y, groups=df['subject'], cv=LeaveOneGroupOut())\n",
    "test_score(cv_rf_all)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD4CAYAAAD8Zh1EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAATjElEQVR4nO3dfZCdZ3nf8e8vMgpgCJBoISDJSA0KRcMwMdkqpFDKxLiVDZXi2GnlQgenEE07KCYxfRGF0SROO1NChqRM1SSqgTIEozgGpwtRItNg2qYDrtb4BUuKYBEGrcB4IU5oyxQjuPrHedQc1iuds7v3WtLj72dmR8/Lretcs9L57bP383JSVUiSLnzfd64bkCS1YaBLUk8Y6JLUEwa6JPWEgS5JPXHRuXrhNWvW1IYNG87Vy0vSBemuu+76WlVNLLTvnAX6hg0bmJ6ePlcvL0kXpCRfPNM+p1wkqScMdEnqCQNdknrCQJeknjDQJaknDHRJ6gkDXZJ6wkCXpJ4w0CWpJ87ZnaIXsk/ue3WTOj+586NN6kgSeIQuSb0xVqAn2ZrkWJKZJLsX2H9JkjuS3J3kviRXtm9VknQ2IwM9ySpgL3AFsBm4NsnmecPeBtxSVZcCO4D/0LpRSdLZjXOEvgWYqarjVfUIsB/YPm9MAT/QLT8N+HK7FiVJ4xgn0NcCJ4bWZ7ttw34ZeG2SWeAA8AsLFUqyM8l0kum5ubkltCtJOpNWJ0WvBf5TVa0DrgTen+RRtatqX1VNVtXkxMSCz2eXJC3ROIF+Elg/tL6u2zbs9cAtAFX1SeCJwJoWDUqSxjNOoB8CNiXZmGQ1g5OeU/PGfAm4DCDJCxgEunMqkvQYGhnoVXUK2AUcBI4yuJrlcJIbk2zrhr0Z+Pkk9wIfBK6rqlqppiVJjzbWnaJVdYDByc7hbXuGlo8AL23bmiRpMbxTVJJ6wkCXpJ4w0CWpJwx0SeoJA12SesJAl6SeMNAlqScMdEnqCQNdknrCQJeknjDQJaknDHRJ6gkDXZJ6wkCXpJ4w0CWpJwx0SeqJsQI9ydYkx5LMJNm9wP7fSHJP9/XZJH/RvlVJ0tmM/MSiJKuAvcDlwCxwKMlU9ylFAFTVLw2N/wXg0hXoVZJ0FuMcoW8BZqrqeFU9AuwHtp9l/LUMPldUkvQYGifQ1wInhtZnu22PkuS5wEbg42fYvzPJdJLpubm5xfYqSTqL1idFdwC3VtV3FtpZVfuqarKqJicmJhq/tCQ9vo0T6CeB9UPr67ptC9mB0y2SdE6ME+iHgE1JNiZZzSC0p+YPSvLXgWcAn2zboiRpHCMDvapOAbuAg8BR4JaqOpzkxiTbhobuAPZXVa1Mq5Kksxl52SJAVR0ADszbtmfe+i+3a0uStFhjBboeG7e+d2uTOtf83B83qSPpwmKga1l+7rY2P4Tee5U/hKTl8lkuktQTBrok9YSBLkk9YaBLUk8Y6JLUEwa6JPWEgS5JPWGgS1JPGOiS1BMGuiT1hIEuST1hoEtST/hwLknNffjWrzWp8zPXrGlS5/HCI3RJ6omxAj3J1iTHkswk2X2GMX8/yZEkh5Pc3LZNSdIoI6dckqwC9gKXA7PAoSRTVXVkaMwm4C3AS6vq4STPXKmGJUkLG+cIfQswU1XHq+oRYD+wfd6Ynwf2VtXDAFX1UNs2JUmjjBPoa4ETQ+uz3bZhPwr8aJL/keRTSRb8GJskO5NMJ5mem5tbWseSpAW1Oil6EbAJeAVwLfAfkzx9/qCq2ldVk1U1OTEx0eilJUkw3mWLJ4H1Q+vrum3DZoE7q+rbwBeSfJZBwB9q0qWW7V0f+LvLrnH9aw426ETSShnnCP0QsCnJxiSrgR3A1Lwxf8Dg6JwkaxhMwRxv2KckaYSRgV5Vp4BdwEHgKHBLVR1OcmOSbd2wg8DXkxwB7gD+eVV9faWaliQ92lh3ilbVAeDAvG17hpYLuKH7kiSdA94pKkk94bNcpMex6287MXrQCO+6av3oQXpMeIQuST1hoEtSTxjoktQTzqHrvHTlbf+6SZ0DV72tSR3pQuARuiT1hIEuST1hoEtSTxjoktQTnhSVznNXfehPm9S57eqXNamj85dH6JLUEwa6JPWEgS5JPXHO59Dnfut3m9SZ+KevbVJHki5UHqFLUk+MdYSeZCvw74BVwE1V9W/n7b8OeAd/9Vmj/76qbmrYp9TMqz60/P+af3j1Gxp0IrU1MtCTrAL2Apcz+DDoQ0mmqurIvKG/V1W7VqBHSdIYxply2QLMVNXxqnoE2A9sX9m2JEmLNU6grwWGP9Zktts239VJ7ktya5IFP8Ikyc4k00mm5+bmltCuJOlMWp0U/QiwoapeBHwMeN9Cg6pqX1VNVtXkxMREo5eWJMF4gX4SGD7iXsdfnfwEoKq+XlXf6lZvAn68TXuSpHGNc5XLIWBTko0MgnwH8A+HByR5dlV9pVvdBhxt2uUSfXnvDcuu8Zw3vrNBJ5K08kYGelWdSrILOMjgssX3VNXhJDcC01U1BVyfZBtwCvhz4LoV7FmStICxrkOvqgPAgXnb9gwtvwV4S9vWJEmL4Z2iktQTBrok9YSBLkk9YaBLUk8Y6JLUEwa6JPWEgS5JPWGgS1JPGOiS1BMGuiT1hIEuST1hoEtSTxjoktQTBrok9YSBLkk9YaBLUk8Y6JLUE2MFepKtSY4lmUmy+yzjrk5SSSbbtShJGsfIQE+yCtgLXAFsBq5NsnmBcU8F3gTc2bpJSdJo4xyhbwFmqup4VT0C7Ae2LzDuV4G3A/+3YX+SpDGNE+hrgRND67Pdtv8vyYuB9VX1h2crlGRnkukk03Nzc4tuVpJ0Zss+KZrk+4B3Am8eNbaq9lXVZFVNTkxMLPelJUlDxgn0k8D6ofV13bbTngq8EPhEkgeAlwBTnhiVpMfWOIF+CNiUZGOS1cAOYOr0zqr6y6paU1UbqmoD8ClgW1VNr0jHkqQFjQz0qjoF7AIOAkeBW6rqcJIbk2xb6QYlSeO5aJxBVXUAODBv254zjH3F8tuSJC2Wd4pKUk8Y6JLUEwa6JPWEgS5JPWGgS1JPGOiS1BMGuiT1xFjXoUuSluar7/pEkzrPuv4VI8d4hC5JPWGgS1JPGOiS1BMGuiT1hIEuST1hoEtSTxjoktQTBrok9cRYgZ5ka5JjSWaS7F5g/z9J8pkk9yT50ySb27cqSTqbkXeKJlkF7AUuB2aBQ0mmqurI0LCbq+q3u/HbgHcCW1egX0mPY3ff9FCTOpe+4ZlN6pxvxjlC3wLMVNXxqnoE2A9sHx5QVd8YWr0YqHYtSpLGMc6zXNYCJ4bWZ4GfmD8oyRuBG4DVwE8tVCjJTmAnwCWXXLLYXiVJZ9HspGhV7a2qHwH+JfC2M4zZV1WTVTU5MTHR6qUlSYwX6CeB9UPr67ptZ7If+OnlNCVJWrxxAv0QsCnJxiSrgR3A1PCAJJuGVl8FfK5di5KkcYycQ6+qU0l2AQeBVcB7qupwkhuB6aqaAnYleSXwbeBh4HUr2bQk6dHG+oCLqjoAHJi3bc/Q8psa9yVJWiTvFJWknjDQJaknDHRJ6gkDXZJ6YqyTopLUZw/++kyTOj/8z57XpM5SeYQuST1hoEtSTxjoktQTBrok9YSBLkk9YaBLUk8Y6JLUEwa6JPWEgS5JPWGgS1JPGOiS1BNjBXqSrUmOJZlJsnuB/TckOZLkviR/kuS57VuVJJ3NyEBPsgrYC1wBbAauTbJ53rC7gcmqehFwK/BrrRuVJJ3dOEfoW4CZqjpeVY8A+4HtwwOq6o6q+ma3+ilgXds2JUmjjBPoa4ETQ+uz3bYzeT3wRwvtSLIzyXSS6bm5ufG7lCSN1PSkaJLXApPAOxbaX1X7qmqyqiYnJiZavrQkPe6N8wEXJ4H1Q+vrum3fI8krgbcCf7uqvtWmPUnSuMY5Qj8EbEqyMclqYAcwNTwgyaXA7wDbquqh9m1KkkYZGehVdQrYBRwEjgK3VNXhJDcm2dYNewfwFOD3k9yTZOoM5SRJK2SszxStqgPAgXnb9gwtv7JxX5KkRfJOUUnqCQNdknrCQJeknjDQJaknDHRJ6gkDXZJ6wkCXpJ4w0CWpJwx0SeoJA12SesJAl6SeMNAlqScMdEnqCQNdknrCQJeknjDQJaknxgr0JFuTHEsyk2T3AvtfnuTTSU4luaZ9m5KkUUYGepJVwF7gCmAzcG2SzfOGfQm4Dri5dYOSpPGM8xF0W4CZqjoOkGQ/sB04cnpAVT3Q7fvuCvQoSRrDOFMua4ETQ+uz3TZJ0nnkMT0pmmRnkukk03Nzc4/lS0tS740T6CeB9UPr67pti1ZV+6pqsqomJyYmllJCknQG4wT6IWBTko1JVgM7gKmVbUuStFgjA72qTgG7gIPAUeCWqjqc5MYk2wCS/I0ks8DPAr+T5PBKNi1JerRxrnKhqg4AB+Zt2zO0fIjBVIwk6RzxTlFJ6gkDXZJ6wkCXpJ4w0CWpJwx0SeoJA12SesJAl6SeMNAlqScMdEnqCQNdknrCQJeknjDQJaknDHRJ6gkDXZJ6wkCXpJ4w0CWpJwx0SeqJsQI9ydYkx5LMJNm9wP7vT/J73f47k2xo3agk6exGfgRdklXAXuByYBY4lGSqqo4MDXs98HBVPS/JDuDtwD9YiYal89Xfu/XDTep85JqfaVJHjz/jHKFvAWaq6nhVPQLsB7bPG7MdeF+3fCtwWZK0a1OSNEqq6uwDkmuArVX1hm79HwE/UVW7hsbc342Z7dY/34352rxaO4Gd3erzgWNj9rkG+NrIUUuzUrUvtLorWdu6K1/7Qqu7krUvtLqLrf3cqppYaMfIKZeWqmofsG+xfy/JdFVNrkBLK1b7Qqu7krWtu/K1L7S6K1n7QqvbsvY4Uy4ngfVD6+u6bQuOSXIR8DTg68ttTpI0vnEC/RCwKcnGJKuBHcDUvDFTwOu65WuAj9eouRxJUlMjp1yq6lSSXcBBYBXwnqo6nORGYLqqpoB3A+9PMgP8OYPQb2nR0zTnQe0Lre5K1rbuyte+0OquZO0LrW6z2iNPikqSLgzeKSpJPWGgS1JPnPeBPuqxA0us+Z4kD3XXzzeTZH2SO5IcSXI4yZsa1n5ikv+Z5N6u9q+0qt3VX5Xk7iQfbVjzgSSfSXJPkulWdbvaT09ya5I/S3I0yU82qPn8rtfTX99I8ouN+v2l7t/t/iQfTPLEFnW72m/q6h5eTr8LvS+S/GCSjyX5XPfnMxrW/tmu5+8mWdIle2eo+47u/8V9SW5L8vRGdX+1q3lPktuTPKdVz0P73pykkqxZSm2q6rz9YnAS9vPAXwNWA/cCmxvUfTnwYuD+xv0+G3hxt/xU4LMt+u3qBXhKt/wE4E7gJQ17vwG4Gfhow5oPAGtW6P/G+4A3dMurgac3rr8KeJDBTRzLrbUW+ALwpG79FuC6Rn2+ELgfeDKDixz+C/C8JdZ61PsC+DVgd7e8G3h7w9ovYHCD4SeAyYZ1/w5wUbf89qX0fIa6PzC0fD3w26167ravZ3DxyReX+r4534/Qx3nswKJV1X9jcDVOU1X1lar6dLf8v4CjDN7MLWpXVf3vbvUJ3VeTM9pJ1gGvAm5qUW+lJXkagzfFuwGq6pGq+ovGL3MZ8Pmq+mKjehcBT+ru03gy8OVGdV8A3FlV36yqU8B/BZb0MJgzvC+GH+vxPuCnW9WuqqNVNe7d4oupe3v3vQD4FIN7Z1rU/cbQ6sUs8f13lvz5DeBfLLUunP9TLmuBE0PrszQKyJXWPXHyUgZH0q1qrkpyD/AQ8LGqalX7Nxn8R/puo3qnFXB7kru6xz60shGYA97bTRPdlOTihvVhcOntB1sUqqqTwK8DXwK+AvxlVd3eojaDo/O/leSHkjwZuJLvvRFwuZ5VVV/plh8EntWw9mPhHwN/1KpYkn+T5ATwGmBPw7rbgZNVde9y6pzvgX5BSvIU4EPAL877qb4sVfWdqvoxBkccW5K8cLk1k7waeKiq7lp2g4/2sqp6MXAF8MYkL29U9yIGv7L+VlVdCvwfBtMBTXQ30G0Dfr9RvWcwONLdCDwHuDjJa1vUrqqjDKYVbgf+GLgH+E6L2gu8VtHot8LHQpK3AqeAD7SqWVVvrar1Xc1do8aPo/tB/K9o8APifA/0cR47cF5J8gQGYf6BqmrzPNV5uumFO4CtDcq9FNiW5AEGU1o/leR3G9Q9fWRKVT0E3MZgCq2FWWB26DeUWxkEfCtXAJ+uqq82qvdK4AtVNVdV3wY+DPzNRrWpqndX1Y9X1cuBhxmcu2nlq0meDdD9+VDD2ismyXXAq4HXdD+IWvsAcHWjWj/C4If9vd37cB3w6SQ/vNhC53ugj/PYgfNGkjCY1z1aVe9sXHvi9Nn6JE9i8Hz6P1tu3ap6S1Wtq6oNDL6/H6+qZR89Jrk4yVNPLzM4UdXkqqKqehA4keT53abLgCNn+SuLdS2Npls6XwJekuTJ3f+RyxicX2kiyTO7Py9hMH9+c6vafO9jPV4H/OeGtVdEkq0MphC3VdU3G9bdNLS6nQbvP4Cq+kxVPbOqNnTvw1kGF1c8uJRi5/UXgznBzzK42uWtjWp+kMFc5re7b97rG9V9GYNfSe9j8KvvPcCVjWq/CLi7q30/sGcFvtevoNFVLgyuTLq3+zrc6t9uqP6PAdPd9+MPgGc0qnsxgwfLPa1xv7/CIADuB94PfH/D2v+dwQ+0e4HLllHnUe8L4IeAPwE+x+AKmh9sWPuqbvlbwFeBg43qzjA493b6Pbjoq1HOUPdD3b/ffcBHgLWtvhfz9j/AEq9y8dZ/SeqJ833KRZI0JgNdknrCQJeknjDQJaknDHRJ6gkDXZJ6wkCXpJ74f+58Wd1mWHZVAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=list(range(len(subjects))), y=cv_rf_all['test_accuracy']);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2 0.8196701282834453\n",
      "8 0.005316455696202532\n"
     ]
    }
   ],
   "source": [
    "min_max(cv_rf_all['test_accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Stress vs Non-Stress"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'accuracy': '0.6427685858944838 +-0.14943073781633812', 'f1_macro': '0.5351300574621347 +-0.17146963445978558'}\n"
     ]
    }
   ],
   "source": [
    "clf_b_rf_all = create_rf_pipeline(n_estimators=n_estimators, min_samples_split=min_samples_split,\n",
    "                              criterion=criterion, n_jobs=-1)\n",
    "cv_b_rf_all = run_cv(clf_b_rf_all, X, y_b, groups=df['subject'], cv=LeaveOneGroupOut())\n",
    "print(test_score(cv_b_rf_all))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=list(range(len(subjects))), y=cv_b_rf_all['test_accuracy']);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2 0.8707391569945021\n",
      "13 0.33526367668802365\n"
     ]
    }
   ],
   "source": [
    "min_max(cv_b_rf_all['test_accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {},
   "outputs": [],
   "source": [
    "rf_acc, rf_f1 = raw_scores(cv_rf_all)\n",
    "rf_b_acc, rf_b_f1 = raw_scores(cv_b_rf_all)\n",
    "df_report.loc[modality, 'rf_f1'] = rf_f1\n",
    "df_report.loc[modality, 'rf_acc'] = rf_acc\n",
    "df_report.loc[modality, 'rf_b_f1'] = rf_b_f1\n",
    "df_report.loc[modality, 'rf_b_acc'] = rf_b_acc"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# AdaBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_ab_pipeline(n_estimators=100, min_samples_split=20, criterion='entropy'):\n",
    "    base_estimator = DecisionTreeClassifier(criterion=criterion, min_samples_split=min_samples_split)\n",
    "    model_ab = AdaBoostClassifier(base_estimator=base_estimator, n_estimators=n_estimators)\n",
    "    return make_pipeline(StandardScaler(), model_ab)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Baseline vs Stress vs Amusement"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'accuracy': '0.4947611645544441 +-0.17061247031579288',\n",
       " 'f1_macro': '0.386726902486435 +-0.15451625574115885'}"
      ]
     },
     "execution_count": 206,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_ab_all = create_ab_pipeline()\n",
    "cv_ab_all = run_cv(clf_ab_all, X, y, groups=df['subject'], cv=LeaveOneGroupOut())\n",
    "test_score(cv_ab_all)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=list(range(len(subjects))), y=cv_ab_all['test_accuracy']);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2 0.7997556505803298\n",
      "8 0.12974683544303797\n"
     ]
    }
   ],
   "source": [
    "min_max(cv_ab_all['test_accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Stress vs Non-stress"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'accuracy': '0.6147497201006527 +-0.15978252785041336',\n",
       " 'f1_macro': '0.5004780403703463 +-0.15670744402557127'}"
      ]
     },
     "execution_count": 209,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_b_ab_all = create_ab_pipeline()\n",
    "cv_b_ab_all = run_cv(clf_b_ab_all, X, y_b, groups=df['subject'], cv=LeaveOneGroupOut())\n",
    "test_score(cv_b_ab_all)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=list(range(len(subjects))), y=cv_b_ab_all['test_accuracy']);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2 0.8467929138668295\n",
      "13 0.23410547067520945\n"
     ]
    }
   ],
   "source": [
    "min_max(cv_b_ab_all['test_accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "metadata": {},
   "outputs": [],
   "source": [
    "ab_acc, ab_f1 = raw_scores(cv_ab_all)\n",
    "ab_b_acc, ab_b_f1 = raw_scores(cv_b_ab_all)\n",
    "df_report.loc[modality, 'ab_f1'] = ab_f1\n",
    "df_report.loc[modality, 'ab_acc'] = ab_acc\n",
    "df_report.loc[modality, 'ab_b_f1'] = ab_b_f1\n",
    "df_report.loc[modality, 'ab_b_acc'] = ab_b_acc"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_lda_pipeline():\n",
    "    return make_pipeline(StandardScaler(), LDA())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Baseline vs Stress vs Non-stress"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'accuracy': '0.5754434914420609 +-0.06968548217118739',\n",
       " 'f1_macro': '0.2897186265347211 +-0.10195462544511139'}"
      ]
     },
     "execution_count": 214,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_lda_all = create_lda_pipeline()\n",
    "cv_lda_all = run_cv(model_lda_all, X, y, groups=df['subject'], cv=LeaveOneGroupOut())\n",
    "test_score(cv_lda_all)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=list(range(len(subjects))), y=cv_lda_all['test_accuracy']);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 0.7867191225560324\n",
      "7 0.49806800618238023\n"
     ]
    }
   ],
   "source": [
    "min_max(cv_lda_all['test_accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Stress vs Non-Stress"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'accuracy': '0.7052835593936336 +-0.03168774005457038',\n",
       " 'f1_macro': '0.43031788404186333 +-0.07263065593746956'}"
      ]
     },
     "execution_count": 217,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_b_lda_all = create_lda_pipeline()\n",
    "cv_b_lda_all = run_cv(model_b_lda_all, X, y_b, groups=df['subject'], cv=LeaveOneGroupOut())\n",
    "test_score(cv_b_lda_all)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=list(range(len(subjects))), y=cv_b_lda_all['test_accuracy']);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8 0.809873417721519\n",
      "7 0.6531427099433282\n"
     ]
    }
   ],
   "source": [
    "min_max(cv_b_lda_all['test_accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "metadata": {},
   "outputs": [],
   "source": [
    "lda_acc, lda_f1 = raw_scores(cv_lda_all)\n",
    "lda_b_acc, lda_b_f1 = raw_scores(cv_b_lda_all)\n",
    "df_report.loc[modality, 'lda_f1'] = lda_f1\n",
    "df_report.loc[modality, 'lda_acc'] = lda_acc\n",
    "df_report.loc[modality, 'lda_b_f1'] = lda_b_f1\n",
    "df_report.loc[modality, 'lda_b_acc'] = lda_b_acc"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "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>rf_f1</th>\n",
       "      <th>rf_acc</th>\n",
       "      <th>rf_b_f1</th>\n",
       "      <th>rf_b_acc</th>\n",
       "      <th>ab_f1</th>\n",
       "      <th>ab_acc</th>\n",
       "      <th>ab_b_f1</th>\n",
       "      <th>ab_b_acc</th>\n",
       "      <th>lda_f1</th>\n",
       "      <th>lda_acc</th>\n",
       "      <th>lda_b_f1</th>\n",
       "      <th>lda_b_acc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>ACC</th>\n",
       "      <td>0.395361</td>\n",
       "      <td>0.559278</td>\n",
       "      <td>0.569841</td>\n",
       "      <td>0.723911</td>\n",
       "      <td>0.448804</td>\n",
       "      <td>0.591598</td>\n",
       "      <td>0.588321</td>\n",
       "      <td>0.733027</td>\n",
       "      <td>0.400382</td>\n",
       "      <td>0.597001</td>\n",
       "      <td>0.542222</td>\n",
       "      <td>0.700070</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ECG</th>\n",
       "      <td>0.518569</td>\n",
       "      <td>0.660007</td>\n",
       "      <td>0.796300</td>\n",
       "      <td>0.839966</td>\n",
       "      <td>0.515433</td>\n",
       "      <td>0.662472</td>\n",
       "      <td>0.800814</td>\n",
       "      <td>0.844100</td>\n",
       "      <td>0.515813</td>\n",
       "      <td>0.708160</td>\n",
       "      <td>0.809485</td>\n",
       "      <td>0.865115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EDA</th>\n",
       "      <td>0.643974</td>\n",
       "      <td>0.667019</td>\n",
       "      <td>0.749781</td>\n",
       "      <td>0.787745</td>\n",
       "      <td>0.621201</td>\n",
       "      <td>0.645761</td>\n",
       "      <td>0.741442</td>\n",
       "      <td>0.781814</td>\n",
       "      <td>0.451855</td>\n",
       "      <td>0.574640</td>\n",
       "      <td>0.692621</td>\n",
       "      <td>0.764252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EMG</th>\n",
       "      <td>0.462594</td>\n",
       "      <td>0.607954</td>\n",
       "      <td>0.599067</td>\n",
       "      <td>0.713775</td>\n",
       "      <td>0.480842</td>\n",
       "      <td>0.620192</td>\n",
       "      <td>0.610554</td>\n",
       "      <td>0.723836</td>\n",
       "      <td>0.445108</td>\n",
       "      <td>0.599009</td>\n",
       "      <td>0.614172</td>\n",
       "      <td>0.724727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RESP</th>\n",
       "      <td>0.545764</td>\n",
       "      <td>0.676278</td>\n",
       "      <td>0.786080</td>\n",
       "      <td>0.824339</td>\n",
       "      <td>0.544442</td>\n",
       "      <td>0.682530</td>\n",
       "      <td>0.785471</td>\n",
       "      <td>0.821546</td>\n",
       "      <td>0.545973</td>\n",
       "      <td>0.698706</td>\n",
       "      <td>0.796305</td>\n",
       "      <td>0.841577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TEMP</th>\n",
       "      <td>0.408837</td>\n",
       "      <td>0.499755</td>\n",
       "      <td>0.535130</td>\n",
       "      <td>0.642769</td>\n",
       "      <td>0.386727</td>\n",
       "      <td>0.494761</td>\n",
       "      <td>0.500478</td>\n",
       "      <td>0.614750</td>\n",
       "      <td>0.289719</td>\n",
       "      <td>0.575443</td>\n",
       "      <td>0.430318</td>\n",
       "      <td>0.705284</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         rf_f1    rf_acc   rf_b_f1  rf_b_acc     ab_f1    ab_acc   ab_b_f1  \\\n",
       "ACC   0.395361  0.559278  0.569841  0.723911  0.448804  0.591598  0.588321   \n",
       "ECG   0.518569  0.660007  0.796300  0.839966  0.515433  0.662472  0.800814   \n",
       "EDA   0.643974  0.667019  0.749781  0.787745  0.621201  0.645761  0.741442   \n",
       "EMG   0.462594  0.607954  0.599067  0.713775  0.480842  0.620192  0.610554   \n",
       "RESP  0.545764  0.676278  0.786080  0.824339  0.544442  0.682530  0.785471   \n",
       "TEMP  0.408837  0.499755  0.535130  0.642769  0.386727  0.494761  0.500478   \n",
       "\n",
       "      ab_b_acc    lda_f1   lda_acc  lda_b_f1  lda_b_acc  \n",
       "ACC   0.733027  0.400382  0.597001  0.542222   0.700070  \n",
       "ECG   0.844100  0.515813  0.708160  0.809485   0.865115  \n",
       "EDA   0.781814  0.451855  0.574640  0.692621   0.764252  \n",
       "EMG   0.723836  0.445108  0.599009  0.614172   0.724727  \n",
       "RESP  0.821546  0.545973  0.698706  0.796305   0.841577  \n",
       "TEMP  0.614750  0.289719  0.575443  0.430318   0.705284  "
      ]
     },
     "execution_count": 221,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_report.to_csv('chest_scores_per_mod.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df_report.to_markdown()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}