notebooks/regression_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from scipy.stats import stats\n",
    "import os\n",
    "import matplotlib.pyplot as plt\n",
    "from random import choices\n",
    "import xgboost as xgb\n",
    "\n",
    "from sklearn.model_selection import GridSearchCV, train_test_split\n",
    "\n",
    "from sklearn.model_selection import StratifiedKFold, KFold\n",
    "\n",
    "from sklearn import metrics\n",
    "from sklearn.kernel_ridge import KernelRidge\n",
    "from sklearn.linear_model import Lasso, LassoCV, ElasticNet, ElasticNetCV\n",
    "from sklearn.ensemble import RandomForestRegressor, GradientBoostingRegressor\n",
    "from sklearn.metrics import mean_squared_error, mean_absolute_percentage_error, r2_score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Filter_low_Expr(arr):\n",
    "    total_len = len(arr)\n",
    "    count = 0\n",
    "    for i in arr:\n",
    "        if i > 0:\n",
    "            count = count + 1\n",
    "    if count > 0.5*total_len: #If \n",
    "        return(True)\n",
    "    else:\n",
    "        return(False)\n",
    "    \n",
    "def formating_relative_expression(input_data, selected_genes_id, sele_house_keeping_id, KG_reg, dic_gene_map):\n",
    "    array_list = []\n",
    "    new_id_list = []\n",
    "    print(\"House keeping genes:\")\n",
    "    for gene1 in selected_genes_id:\n",
    "\n",
    "        for gene2 in sele_house_keeping_id:\n",
    "            #print(gene2 + \":\" + gene1)\n",
    "            new_arr = (input_data.loc[[gene1],:].values - input_data.loc[[gene2],:].values)[0]\n",
    "            new_arr_bi = []\n",
    "            for i in new_arr:\n",
    "                if i > 0:\n",
    "                    new_arr_bi.append(1)\n",
    "                else:\n",
    "                    new_arr_bi.append(0)\n",
    "            array_list.append(new_arr_bi)\n",
    "            new_id_list.append(dic_gene_map[gene1] + '-' + dic_gene_map[gene2])\n",
    "\n",
    "    print(\"Regulators:\")\n",
    "    for gene1 in selected_genes_id:\n",
    "        cur_sele = KG_reg.loc[KG_reg['Target_gene'] == gene1]\n",
    "        cur_regulators = list(cur_sele.loc[cur_sele['Coef'] < 0]['Regulator'])\n",
    "        if gene1+\"_intercept_\" in cur_regulators:\n",
    "            cur_regulators.remove(gene1+\"_intercept_\")\n",
    "        #cur_regulators.remove(gene1+\"_RSS\")\n",
    "        cur_regulators_ids = []\n",
    "        for gene2 in cur_regulators:\n",
    "            gene2 = dic_gene_map[gene2]\n",
    "            if gene2 in dic_gene_map and gene2 in input_data.index:\n",
    "                #print(gene2 + \":\" + gene1)\n",
    "                \n",
    "                #cur_regulators_ids.append(dic_gene_map[i])\n",
    "                new_arr = (input_data.loc[[gene1],:].values - input_data.loc[[gene2],:].values)[0]\n",
    "                new_arr_bi = []\n",
    "                for i in new_arr:\n",
    "                    if i > 0:\n",
    "                        new_arr_bi.append(1)\n",
    "                    else:\n",
    "                        new_arr_bi.append(0)\n",
    "                array_list.append(new_arr_bi)\n",
    "                new_id_list.append(dic_gene_map[gene1] + '-' + dic_gene_map[gene2])\n",
    "\n",
    "    new_feature_df = pd.DataFrame(array_list)\n",
    "    new_feature_df.columns = input_data.columns\n",
    "    new_feature_df.index = new_id_list\n",
    "    new_feature_df_t = new_feature_df.transpose()\n",
    "    return(new_feature_df_t)\n",
    "\n",
    "\n",
    "def train_test_split_gq(df_sele, binary_cutoff, test_size):\n",
    "    from sklearn.model_selection import train_test_split\n",
    "    df_neg = df_sele.loc[df_sele['ic50'] > binary_cutoff]\n",
    "    df_train_1, df_test_1 = train_test_split(df_neg, test_size = test_size)\n",
    "    df_pos = df_sele.loc[df_sele['ic50'] <= binary_cutoff]\n",
    "    df_train_2, df_test_2 = train_test_split(df_pos, test_size = test_size)\n",
    "    \n",
    "    df_train = pd.concat([df_train_1,df_train_2])\n",
    "    df_test = pd.concat([df_test_1, df_test_2])\n",
    "    \n",
    "    return(df_train, df_test)\n",
    "\n",
    "def model_run_with_Independent_testSet(train_data,test_data, train_labels, test_labels, independent_labels, independent_features,\n",
    "                                      model=None, params=None): #modified at 11:46 AM Feb 15\n",
    "    # TODO: convert this to be regression instead of classification\n",
    "    if model is None:\n",
    "        model = xgb.XGBRegressor\n",
    "        if params is None:\n",
    "            params = {'n_jobs': 4, 'learning_rate': 0.05, 'max_depth': 8, 'n_estimators': 100}\n",
    "    else:\n",
    "        if params is None:\n",
    "            params = {}\n",
    "    xgb_model = model(**params)\n",
    "    \n",
    "    xgb_model.fit(train_data,train_labels.values)\n",
    "    \n",
    "    predictions_train = xgb_model.predict(train_data)\n",
    "\n",
    "    predictions = xgb_model.predict(test_data)\n",
    "    \n",
    "    predictions_independent = xgb_model.predict(independent_features)\n",
    "\n",
    "    # evaluation metrics: mean_squared_error, mean_absolute_percentage_error, r2_score\n",
    "\n",
    "    mse_train = mean_squared_error(train_labels, predictions_train)\n",
    "    mape_train = mean_absolute_percentage_error(train_labels, predictions_train)\n",
    "    r2_train = r2_score(train_labels, predictions_train)\n",
    "\n",
    "    # Evaluate the accuracy of the XGBoost model in the validation set (beatAML wave 1/2)\n",
    "    mse_test = mean_squared_error(test_labels, predictions)\n",
    "    mape_test = mean_absolute_percentage_error(test_labels, predictions)\n",
    "    r2_test = r2_score(test_labels, predictions)\n",
    "    \n",
    "    # Evaluate the accuracy of the XGBoost model in the testing set (beatAML wave 3/4)\n",
    "    mse_independent = mean_squared_error(independent_labels, predictions_independent)\n",
    "    mape_independent = mean_absolute_percentage_error(independent_labels, predictions_independent)\n",
    "    r2_independent = r2_score(independent_labels, predictions_independent)\n",
    "\n",
    "    result_summary = {}\n",
    "    #result_summary['random_state'] = random_\n",
    "\n",
    "    result_summary['train_mse'] = mse_train\n",
    "    result_summary['train_mape'] = mape_train\n",
    "    result_summary['train_r2'] = r2_train\n",
    "\n",
    "    #result_summary['best_params'] = clf.best_params_\n",
    "    result_summary['test_mse'] = mse_test #1\n",
    "    result_summary['test_mape'] = mape_test #2\n",
    "    result_summary['test_r2'] = r2_test #3\n",
    "    \n",
    "    result_summary['independent_mse'] = mse_independent #1\n",
    "    result_summary['independent_mape'] = mape_independent #2\n",
    "    result_summary['independent_r2'] = r2_independent #3\n",
    "\n",
    "    if hasattr(xgb_model, 'feature_importances_'):\n",
    "        result_feature = pd.DataFrame({\"Feature\":list(test_data.columns), \"IMP\":xgb_model.feature_importances_})\n",
    "        result_feature = result_feature.loc[result_feature['IMP']>0].sort_values(by = ['IMP'], ascending = False)\n",
    "    else:\n",
    "        result_feature = None\n",
    "    result_summary['features'] = result_feature\n",
    "    return(result_summary)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The BeatAML wave 1/2 dataset is from the Tyler 2018 paper, and the BeatAML wave 3/4 dataset is from the Cancer Cell 2022 paper. The gene expression values are in general similar, but not exactly the same according to the data processing steps.\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "wave 1 to 4 data:\n",
      "(969, 20215)\n",
      "Index(['FLT3', 'SRSF2', 'CEBPA', 'SF3B1', 'NPM1', 'KRAS', 'RUNX1', 'DNMT3A',\n",
      "       'ASXL1', 'IDH2', 'PTPN11', 'JAK2', 'PDS5B', 'STAG2', 'IDH1', 'BCOR',\n",
      "       'TET2', 'GATA2', 'U2AF1', 'EZH2', 'WT1', 'NRAS', 'TP53'],\n",
      "      dtype='object')\n",
      "Index(['DNMT3A:p.R882', 'ASXL1:p.G645-646', 'FLT3:p.835-839', 'IDH1:p.R132',\n",
      "       'IDH2:p.R140', 'IDH2:p.R172', 'JAK2:p.V617', 'KIT:p.D816',\n",
      "       'KRAS:p.G12/13', 'NPM1:p.W288-90', 'NRAS:p.G12/13', 'NRAS:p.Q61',\n",
      "       'SF3B1:p.K700E', 'SF3B1:p.K666', 'SRSF2:p.94-95', 'U2AF1:p.S34',\n",
      "       'U2AF1:p.156-157', 'ZNF687:p.R939Pfs*36:151262327'],\n",
      "      dtype='object')\n",
      "wave 1 to 2 data:\n",
      "(783, 17853)\n",
      "Index(['EZH2', 'DNMT3A', 'BCOR', 'SF3B1', 'TET2', 'KRAS', 'U2AF1', 'CEBPA',\n",
      "       'NPM1', 'GATA2', 'PTPN11', 'IDH2', 'NRAS', 'ASXL1', 'RUNX1', 'JAK2',\n",
      "       'FLT3', 'WT1', 'STAG2', 'TP53', 'IDH1', 'SRSF2'],\n",
      "      dtype='object')\n",
      "Index(['DNMT3A:p.R882', 'ASXL1:p.G645-646', 'FLT3:p.835-839', 'IDH1:p.R132',\n",
      "       'IDH2:p.R140', 'IDH2:p.R172', 'JAK2:p.V617', 'KIT:p.D816',\n",
      "       'KRAS:p.G12/13', 'NPM1:p.W288-90', 'NRAS:p.G12/13', 'NRAS:p.Q61',\n",
      "       'SF3B1:p.K700E', 'SF3B1:p.K666', 'SRSF2:p.94-95', 'U2AF1:p.S34',\n",
      "       'U2AF1:p.156-157', 'ZNF687:p.R939Pfs*36:151262327'],\n",
      "      dtype='object')\n"
     ]
    }
   ],
   "source": [
    "#data_all_w12 = pd.read_csv(\"data/beatAML_wave12_feature_ic50_highExpr_Mar19_2024_w12.csv\", index_col= \"Unnamed: 0\")\n",
    "\n",
    "# Load all data for beat AML wave 1 to 4 from the 2022 paper\n",
    "data_all_w14 = pd.read_csv(\"data/beatAML_wave14_feature_ic50_highExpr_Mar19_2024_w1_4.csv\", index_col= \"Unnamed: 0\")\n",
    "print(\"wave 1 to 4 data:\")\n",
    "print(data_all_w14.shape)\n",
    "print(data_all_w14.columns[0:23]) # Feature set 1 is from column 0 to 23\n",
    "print(data_all_w14.columns[23:41]) # Feature set 2 is from column 23 to 41\n",
    "#print(data_all_w14.columns[41:19894]) # Feature set 3 is from column 41 to 19894\n",
    "#print(data_all_w14.columns[19894:20215]) # Drug response data is from column 19894 to 20215\n",
    "\n",
    "# Load all data from the beat AML wave 1 to 2 data from the 2018 paper\n",
    "data_all_w12 = pd.read_csv(\"data/beatAML_wave12_feature_ic50_highExpr_Apr30_2024_w12.csv\", index_col= \"Unnamed: 0\")\n",
    "print(\"wave 1 to 2 data:\")\n",
    "print(data_all_w12.shape)\n",
    "print(data_all_w12.columns[0:22])\n",
    "print(data_all_w12.columns[22:40])\n",
    "#print(data_all_w12.columns[40:17731])\n",
    "#print(data_all_w12.columns[17731:17853])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Theare are 672 samples from the Beat AML wave 1/2 and 297 samples from wave 3/4\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "672\n",
      "297\n"
     ]
    }
   ],
   "source": [
    "w12_shared_samples = list(set(data_all_w12.index).intersection(set(data_all_w14.index)))\n",
    "print(len(w12_shared_samples))\n",
    "\n",
    "w34_samples = list(set(data_all_w14.index) - set(w12_shared_samples))\n",
    "print(len(w34_samples))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(783, 17853)"
      ]
     },
     "execution_count": 211,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_all_w12.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# Load testing data from BeatAML wave 3 and 4\n",
    "BeatAML_wave_3_expr = pd.read_csv(\"data/beataml_waves1to4_allnorm.txt.gz\", sep = \"\\t\") \n",
    "BeatAML_wave_3_drugResponse = pd.read_csv(\"data/beataml_probit_curve_fits_v4_distr_nonproprietary.txt.gz\", sep = \"\\t\")\n",
    "\n",
    "# read xlsx file\n",
    "#BeatAML_all_clinic = pd.read_excel(\"data/beataml_wv1to4_clinical_summary_01_22_2021.xlsx\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['lab_id', 'patient_id', 'inhibitor', 'min_conc', 'max_conc',\n",
       "       'intercept', 'beta', 'beta_z', 'beta_p', 'aic', 'pearson_chisq',\n",
       "       'deviance', 'converged', 'ic10', 'ic25', 'ic50', 'ic75', 'ic90', 'auc',\n",
       "       'all_gt_50', 'all_lt_50', 'curve_type', 'type', 'proprietary'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 213,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "BeatAML_wave_3_drugResponse.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {},
   "outputs": [],
   "source": [
    "BeatAML_wave_3_drugResponse.index = pd.MultiIndex.from_arrays([BeatAML_wave_3_drugResponse.lab_id, BeatAML_wave_3_drugResponse.inhibitor])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Adding AUC data for all drugs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# TODO: add columns to data_all_w14 and data_all_w12 for the drug AUC for all drugs.\n",
    "patient_drugs_auc = {}\n",
    "for i, row in data_all_w14.iterrows():\n",
    "    patient_auc = {}\n",
    "    for drug in data_all_w14.columns[19894:20215]:\n",
    "        try:\n",
    "            patient_auc[drug + '_auc'] = BeatAML_wave_3_drugResponse.loc[(i, drug), 'auc']\n",
    "        except KeyError:\n",
    "            continue\n",
    "    patient_drugs_auc[i] = patient_auc\n",
    "\n",
    "patient_drugs_auc_w12 = {}\n",
    "for i, row in data_all_w12.iterrows():\n",
    "    patient_auc = {}\n",
    "    for drug in data_all_w12.columns[17731:17853]:\n",
    "        try:\n",
    "            patient_auc[drug + '_auc'] = BeatAML_wave_3_drugResponse.loc[(i, drug), 'auc']\n",
    "        except KeyError:\n",
    "            continue\n",
    "    patient_drugs_auc_w12[i] = patient_auc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "metadata": {},
   "outputs": [],
   "source": [
    "patient_drugs_auc = pd.DataFrame(patient_drugs_auc).T\n",
    "patient_drugs_auc_w12 = pd.DataFrame(patient_drugs_auc_w12).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "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>FLT3</th>\n",
       "      <th>SRSF2</th>\n",
       "      <th>CEBPA</th>\n",
       "      <th>SF3B1</th>\n",
       "      <th>NPM1</th>\n",
       "      <th>KRAS</th>\n",
       "      <th>RUNX1</th>\n",
       "      <th>DNMT3A</th>\n",
       "      <th>ASXL1</th>\n",
       "      <th>IDH2</th>\n",
       "      <th>...</th>\n",
       "      <th>PD98059_auc</th>\n",
       "      <th>PD153035_auc</th>\n",
       "      <th>R547_auc</th>\n",
       "      <th>R406_auc</th>\n",
       "      <th>MEK1/2 Inhibitor_auc</th>\n",
       "      <th>Everolimus_auc</th>\n",
       "      <th>AKT Inhibitor X_auc</th>\n",
       "      <th>PRT062070_auc</th>\n",
       "      <th>AST-487_auc</th>\n",
       "      <th>SU14813_auc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>11-00178</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17-00066</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18-00208</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15-00626</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15-00357</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</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",
       "      <td>...</td>\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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13-00316</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16-00541</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16-00524</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12-00301</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14-00640</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>969 rows × 20536 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          FLT3  SRSF2  CEBPA  SF3B1  NPM1  KRAS  RUNX1  DNMT3A  ASXL1  IDH2  \\\n",
       "11-00178   1.0    0.0    0.0    0.0   1.0   0.0    0.0     1.0    0.0   1.0   \n",
       "17-00066   0.0    1.0    0.0    0.0   0.0   0.0    1.0     1.0    0.0   1.0   \n",
       "18-00208   0.0    0.0    0.0    0.0   0.0   0.0    0.0     0.0    0.0   1.0   \n",
       "15-00626   0.0    0.0    0.0    0.0   0.0   0.0    0.0     0.0    0.0   1.0   \n",
       "15-00357   0.0    0.0    0.0    0.0   0.0   0.0    0.0     1.0    0.0   1.0   \n",
       "...        ...    ...    ...    ...   ...   ...    ...     ...    ...   ...   \n",
       "13-00316   NaN    NaN    NaN    NaN   NaN   NaN    NaN     NaN    NaN   NaN   \n",
       "16-00541   NaN    NaN    NaN    NaN   NaN   NaN    NaN     NaN    NaN   NaN   \n",
       "16-00524   NaN    NaN    NaN    NaN   NaN   NaN    NaN     NaN    NaN   NaN   \n",
       "12-00301   NaN    NaN    NaN    NaN   NaN   NaN    NaN     NaN    NaN   NaN   \n",
       "14-00640   NaN    NaN    NaN    NaN   NaN   NaN    NaN     NaN    NaN   NaN   \n",
       "\n",
       "          ...  PD98059_auc  PD153035_auc  R547_auc  R406_auc  \\\n",
       "11-00178  ...          NaN           NaN       NaN       NaN   \n",
       "17-00066  ...          NaN           NaN       NaN       NaN   \n",
       "18-00208  ...          NaN           NaN       NaN       NaN   \n",
       "15-00626  ...          NaN           NaN       NaN       NaN   \n",
       "15-00357  ...          NaN           NaN       NaN       NaN   \n",
       "...       ...          ...           ...       ...       ...   \n",
       "13-00316  ...          NaN           NaN       NaN       NaN   \n",
       "16-00541  ...          NaN           NaN       NaN       NaN   \n",
       "16-00524  ...          NaN           NaN       NaN       NaN   \n",
       "12-00301  ...          NaN           NaN       NaN       NaN   \n",
       "14-00640  ...          NaN           NaN       NaN       NaN   \n",
       "\n",
       "          MEK1/2 Inhibitor_auc  Everolimus_auc  AKT Inhibitor X_auc  \\\n",
       "11-00178                   NaN             NaN                  NaN   \n",
       "17-00066                   NaN             NaN                  NaN   \n",
       "18-00208                   NaN             NaN                  NaN   \n",
       "15-00626                   NaN             NaN                  NaN   \n",
       "15-00357                   NaN             NaN                  NaN   \n",
       "...                        ...             ...                  ...   \n",
       "13-00316                   NaN             NaN                  NaN   \n",
       "16-00541                   NaN             NaN                  NaN   \n",
       "16-00524                   NaN             NaN                  NaN   \n",
       "12-00301                   NaN             NaN                  NaN   \n",
       "14-00640                   NaN             NaN                  NaN   \n",
       "\n",
       "          PRT062070_auc  AST-487_auc  SU14813_auc  \n",
       "11-00178            NaN          NaN          NaN  \n",
       "17-00066            NaN          NaN          NaN  \n",
       "18-00208            NaN          NaN          NaN  \n",
       "15-00626            NaN          NaN          NaN  \n",
       "15-00357            NaN          NaN          NaN  \n",
       "...                 ...          ...          ...  \n",
       "13-00316            NaN          NaN          NaN  \n",
       "16-00541            NaN          NaN          NaN  \n",
       "16-00524            NaN          NaN          NaN  \n",
       "12-00301            NaN          NaN          NaN  \n",
       "14-00640            NaN          NaN          NaN  \n",
       "\n",
       "[969 rows x 20536 columns]"
      ]
     },
     "execution_count": 219,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_all_w14 = pd.concat([data_all_w14, patient_drugs_auc], axis=1)\n",
    "data_all_w14"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "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>EZH2</th>\n",
       "      <th>DNMT3A</th>\n",
       "      <th>BCOR</th>\n",
       "      <th>SF3B1</th>\n",
       "      <th>TET2</th>\n",
       "      <th>KRAS</th>\n",
       "      <th>U2AF1</th>\n",
       "      <th>CEBPA</th>\n",
       "      <th>NPM1</th>\n",
       "      <th>GATA2</th>\n",
       "      <th>...</th>\n",
       "      <th>Entospletinib (GS-9973)_auc</th>\n",
       "      <th>Linifanib (ABT-869)_auc</th>\n",
       "      <th>NVP-ADW742_auc</th>\n",
       "      <th>Lenalidomide_auc</th>\n",
       "      <th>Nutlin 3a_auc</th>\n",
       "      <th>Lenvatinib_auc</th>\n",
       "      <th>TG100-115_auc</th>\n",
       "      <th>Volasertib (BI-6727)_auc</th>\n",
       "      <th>Ruxolitinib (INCB018424)_auc</th>\n",
       "      <th>Crenolanib_auc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>14-00359</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16-00867</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>121.530149</td>\n",
       "      <td>170.212221</td>\n",
       "      <td>105.128902</td>\n",
       "      <td>80.927174</td>\n",
       "      <td>273.178351</td>\n",
       "      <td>124.250679</td>\n",
       "      <td>226.460228</td>\n",
       "      <td>103.654599</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13-00602</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>182.196412</td>\n",
       "      <td>222.667803</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>243.212659</td>\n",
       "      <td>NaN</td>\n",
       "      <td>186.065714</td>\n",
       "      <td>266.898220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15-00302</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>262.600099</td>\n",
       "      <td>210.351498</td>\n",
       "      <td>202.644853</td>\n",
       "      <td>NaN</td>\n",
       "      <td>251.613839</td>\n",
       "      <td>200.337986</td>\n",
       "      <td>249.389303</td>\n",
       "      <td>163.109627</td>\n",
       "      <td>259.536301</td>\n",
       "      <td>243.871620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13-00454</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>226.443072</td>\n",
       "      <td>267.823353</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>286.327943</td>\n",
       "      <td>NaN</td>\n",
       "      <td>286.327943</td>\n",
       "      <td>226.856804</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",
       "      <td>...</td>\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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15-00938</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17-00216</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16-00526</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16-01011</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15-00137</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>783 rows × 17973 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          EZH2  DNMT3A  BCOR  SF3B1  TET2  KRAS  U2AF1  CEBPA  NPM1  GATA2  \\\n",
       "14-00359   0.0     0.0   0.0    0.0   0.0   0.0    0.0    0.0   0.0    0.0   \n",
       "16-00867   0.0     1.0   0.0    0.0   0.0   0.0    0.0    0.0   1.0    0.0   \n",
       "13-00602   1.0     0.0   0.0    0.0   1.0   0.0    0.0    1.0   0.0    0.0   \n",
       "15-00302   0.0     0.0   0.0    0.0   1.0   0.0    0.0    0.0   1.0    0.0   \n",
       "13-00454   0.0     0.0   0.0    0.0   0.0   1.0    0.0    0.0   0.0    0.0   \n",
       "...        ...     ...   ...    ...   ...   ...    ...    ...   ...    ...   \n",
       "15-00938   NaN     NaN   NaN    NaN   NaN   NaN    NaN    NaN   NaN    NaN   \n",
       "17-00216   NaN     NaN   NaN    NaN   NaN   NaN    NaN    NaN   NaN    NaN   \n",
       "16-00526   NaN     NaN   NaN    NaN   NaN   NaN    NaN    NaN   NaN    NaN   \n",
       "16-01011   NaN     NaN   NaN    NaN   NaN   NaN    NaN    NaN   NaN    NaN   \n",
       "15-00137   NaN     NaN   NaN    NaN   NaN   NaN    NaN    NaN   NaN    NaN   \n",
       "\n",
       "          ...  Entospletinib (GS-9973)_auc  Linifanib (ABT-869)_auc  \\\n",
       "14-00359  ...                          NaN                      NaN   \n",
       "16-00867  ...                   121.530149               170.212221   \n",
       "13-00602  ...                          NaN               182.196412   \n",
       "15-00302  ...                   262.600099               210.351498   \n",
       "13-00454  ...                          NaN               226.443072   \n",
       "...       ...                          ...                      ...   \n",
       "15-00938  ...                          NaN                      NaN   \n",
       "17-00216  ...                          NaN                      NaN   \n",
       "16-00526  ...                          NaN                      NaN   \n",
       "16-01011  ...                          NaN                      NaN   \n",
       "15-00137  ...                          NaN                      NaN   \n",
       "\n",
       "          NVP-ADW742_auc  Lenalidomide_auc  Nutlin 3a_auc  Lenvatinib_auc  \\\n",
       "14-00359             NaN               NaN            NaN             NaN   \n",
       "16-00867      105.128902         80.927174     273.178351      124.250679   \n",
       "13-00602      222.667803               NaN            NaN             NaN   \n",
       "15-00302      202.644853               NaN     251.613839      200.337986   \n",
       "13-00454      267.823353               NaN            NaN             NaN   \n",
       "...                  ...               ...            ...             ...   \n",
       "15-00938             NaN               NaN            NaN             NaN   \n",
       "17-00216             NaN               NaN            NaN             NaN   \n",
       "16-00526             NaN               NaN            NaN             NaN   \n",
       "16-01011             NaN               NaN            NaN             NaN   \n",
       "15-00137             NaN               NaN            NaN             NaN   \n",
       "\n",
       "          TG100-115_auc  Volasertib (BI-6727)_auc  \\\n",
       "14-00359            NaN                       NaN   \n",
       "16-00867     226.460228                103.654599   \n",
       "13-00602     243.212659                       NaN   \n",
       "15-00302     249.389303                163.109627   \n",
       "13-00454     286.327943                       NaN   \n",
       "...                 ...                       ...   \n",
       "15-00938            NaN                       NaN   \n",
       "17-00216            NaN                       NaN   \n",
       "16-00526            NaN                       NaN   \n",
       "16-01011            NaN                       NaN   \n",
       "15-00137            NaN                       NaN   \n",
       "\n",
       "          Ruxolitinib (INCB018424)_auc  Crenolanib_auc  \n",
       "14-00359                           NaN             NaN  \n",
       "16-00867                           NaN             NaN  \n",
       "13-00602                    186.065714      266.898220  \n",
       "15-00302                    259.536301      243.871620  \n",
       "13-00454                    286.327943      226.856804  \n",
       "...                                ...             ...  \n",
       "15-00938                           NaN             NaN  \n",
       "17-00216                           NaN             NaN  \n",
       "16-00526                           NaN             NaN  \n",
       "16-01011                           NaN             NaN  \n",
       "15-00137                           NaN             NaN  \n",
       "\n",
       "[783 rows x 17973 columns]"
      ]
     },
     "execution_count": 220,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_all_w12 = pd.concat([data_all_w12, patient_drugs_auc_w12], axis=1)\n",
    "data_all_w12"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Processing sample indices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "672\n"
     ]
    }
   ],
   "source": [
    "BeatAML_w12_samples = set(data_all_w12.index).intersection(set(data_all_w14.index))\n",
    "print(len(BeatAML_w12_samples))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "672\n",
      "297\n"
     ]
    }
   ],
   "source": [
    "BeatAML_w34_samples = set(data_all_w14.index) - BeatAML_w12_samples\n",
    "BeatAML_w12_samples = set(data_all_w14.index).intersection(BeatAML_w12_samples)\n",
    "\n",
    "print(len(BeatAML_w12_samples))\n",
    "print(len(BeatAML_w34_samples))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "metadata": {},
   "outputs": [],
   "source": [
    "BeatAML_w12_data = data_all_w12.loc[list(BeatAML_w12_samples)]  # This is important to make sure which dataset is used for beatAML wave 1/2\n",
    "BeatAML_w34_data = data_all_w14.loc[list(BeatAML_w34_samples)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 224,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(672, 17973)\n",
      "(297, 20536)\n"
     ]
    }
   ],
   "source": [
    "print(BeatAML_w12_data.shape)\n",
    "print(BeatAML_w34_data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[22, 18, 17685]\n"
     ]
    }
   ],
   "source": [
    "BeatAML_w12_data_set1 = BeatAML_w12_data.columns[0:22]\n",
    "BeatAML_w12_data_set2 = BeatAML_w12_data.columns[22:40]\n",
    "BeatAML_w12_data_set3 = BeatAML_w12_data.columns[40:40+17691]\n",
    "\n",
    "\n",
    "BeatAML_w34_data_set1 = BeatAML_w34_data.columns[0:23]\n",
    "BeatAML_w34_data_set2 = BeatAML_w34_data.columns[23:41]\n",
    "BeatAML_w34_data_set3 = BeatAML_w34_data.columns[41:41+19853]\n",
    "\n",
    "featureset_1 = list(set(BeatAML_w12_data_set1).intersection(set(BeatAML_w34_data_set1)))\n",
    "featureset_2 = list(set(BeatAML_w12_data_set2).intersection(set(BeatAML_w34_data_set2)))\n",
    "featureset_3 = list(set(BeatAML_w12_data_set3).intersection(set(BeatAML_w34_data_set3)))\n",
    "\n",
    "print([len(featureset_1), len(featureset_2), len(featureset_3)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 226,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(17685, 672)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>16-00094</th>\n",
       "      <th>16-00533</th>\n",
       "      <th>12-00150</th>\n",
       "      <th>14-00061</th>\n",
       "      <th>14-00476</th>\n",
       "      <th>16-00981</th>\n",
       "      <th>15-00417</th>\n",
       "      <th>14-00473</th>\n",
       "      <th>16-01004</th>\n",
       "      <th>14-00273</th>\n",
       "      <th>...</th>\n",
       "      <th>15-00858</th>\n",
       "      <th>14-00180</th>\n",
       "      <th>13-00625</th>\n",
       "      <th>16-01191</th>\n",
       "      <th>16-01010</th>\n",
       "      <th>15-00909</th>\n",
       "      <th>16-00519</th>\n",
       "      <th>13-00513</th>\n",
       "      <th>14-00542</th>\n",
       "      <th>14-00798</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>ENSG00000116918</th>\n",
       "      <td>5.241544</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.247960</td>\n",
       "      <td>6.022694</td>\n",
       "      <td>5.231849</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.748917</td>\n",
       "      <td>5.151250</td>\n",
       "      <td>5.749562</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>4.093583</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.425932</td>\n",
       "      <td>4.840610</td>\n",
       "      <td>4.001918</td>\n",
       "      <td>5.436100</td>\n",
       "      <td>4.992701</td>\n",
       "      <td>5.397719</td>\n",
       "      <td>4.659582</td>\n",
       "      <td>5.569196</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ENSG00000259943</th>\n",
       "      <td>4.572609</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.230834</td>\n",
       "      <td>3.330238</td>\n",
       "      <td>3.923792</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.059234</td>\n",
       "      <td>4.871109</td>\n",
       "      <td>3.050183</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>4.075204</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.762670</td>\n",
       "      <td>4.290717</td>\n",
       "      <td>4.813993</td>\n",
       "      <td>4.497427</td>\n",
       "      <td>4.146136</td>\n",
       "      <td>3.650255</td>\n",
       "      <td>4.015077</td>\n",
       "      <td>3.761633</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ENSG00000250317</th>\n",
       "      <td>5.171520</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.677739</td>\n",
       "      <td>5.608322</td>\n",
       "      <td>5.159408</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.429578</td>\n",
       "      <td>4.405480</td>\n",
       "      <td>5.407331</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>4.650379</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.305609</td>\n",
       "      <td>4.891720</td>\n",
       "      <td>5.466481</td>\n",
       "      <td>5.059344</td>\n",
       "      <td>4.822493</td>\n",
       "      <td>5.502289</td>\n",
       "      <td>5.244416</td>\n",
       "      <td>4.865866</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ENSG00000271852</th>\n",
       "      <td>1.401104</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.374629</td>\n",
       "      <td>1.397935</td>\n",
       "      <td>1.747647</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.703401</td>\n",
       "      <td>-1.098897</td>\n",
       "      <td>1.619544</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.502477</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.967570</td>\n",
       "      <td>-1.722868</td>\n",
       "      <td>-0.204121</td>\n",
       "      <td>1.756045</td>\n",
       "      <td>0.826824</td>\n",
       "      <td>1.036588</td>\n",
       "      <td>0.066617</td>\n",
       "      <td>-0.181749</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4 rows × 672 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                 16-00094  16-00533  12-00150  14-00061  14-00476  16-00981  \\\n",
       "ENSG00000116918  5.241544       NaN  6.247960  6.022694  5.231849       NaN   \n",
       "ENSG00000259943  4.572609       NaN  4.230834  3.330238  3.923792       NaN   \n",
       "ENSG00000250317  5.171520       NaN  5.677739  5.608322  5.159408       NaN   \n",
       "ENSG00000271852  1.401104       NaN  0.374629  1.397935  1.747647       NaN   \n",
       "\n",
       "                 15-00417  14-00473  16-01004  14-00273  ...  15-00858  \\\n",
       "ENSG00000116918  4.748917  5.151250  5.749562       NaN  ...  4.093583   \n",
       "ENSG00000259943  2.059234  4.871109  3.050183       NaN  ...  4.075204   \n",
       "ENSG00000250317  5.429578  4.405480  5.407331       NaN  ...  4.650379   \n",
       "ENSG00000271852 -0.703401 -1.098897  1.619544       NaN  ... -0.502477   \n",
       "\n",
       "                 14-00180  13-00625  16-01191  16-01010  15-00909  16-00519  \\\n",
       "ENSG00000116918       NaN  4.425932  4.840610  4.001918  5.436100  4.992701   \n",
       "ENSG00000259943       NaN  2.762670  4.290717  4.813993  4.497427  4.146136   \n",
       "ENSG00000250317       NaN  5.305609  4.891720  5.466481  5.059344  4.822493   \n",
       "ENSG00000271852       NaN  1.967570 -1.722868 -0.204121  1.756045  0.826824   \n",
       "\n",
       "                 13-00513  14-00542  14-00798  \n",
       "ENSG00000116918  5.397719  4.659582  5.569196  \n",
       "ENSG00000259943  3.650255  4.015077  3.761633  \n",
       "ENSG00000250317  5.502289  5.244416  4.865866  \n",
       "ENSG00000271852  1.036588  0.066617 -0.181749  \n",
       "\n",
       "[4 rows x 672 columns]"
      ]
     },
     "execution_count": 226,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "BeatAML_w12_data_set3_df = BeatAML_w12_data[featureset_3]\n",
    "BeatAML_w12_data_set3_df_t = BeatAML_w12_data_set3_df.transpose()\n",
    "print(BeatAML_w12_data_set3_df_t.shape)\n",
    "BeatAML_w12_data_set3_df_t.head(4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(17685, 297)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>17-00963</th>\n",
       "      <th>18-00208</th>\n",
       "      <th>17-00256</th>\n",
       "      <th>18-00223</th>\n",
       "      <th>17-00908</th>\n",
       "      <th>17-00838</th>\n",
       "      <th>18-00238</th>\n",
       "      <th>17-00011</th>\n",
       "      <th>17-00048</th>\n",
       "      <th>19-00025</th>\n",
       "      <th>...</th>\n",
       "      <th>00-00015</th>\n",
       "      <th>17-01023</th>\n",
       "      <th>17-01117</th>\n",
       "      <th>00-00004</th>\n",
       "      <th>17-00326</th>\n",
       "      <th>00-00005</th>\n",
       "      <th>17-00368</th>\n",
       "      <th>19-00092</th>\n",
       "      <th>13-00126</th>\n",
       "      <th>17-00857</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>ENSG00000116918</th>\n",
       "      <td>5.527979</td>\n",
       "      <td>5.887300</td>\n",
       "      <td>5.654664</td>\n",
       "      <td>6.607494</td>\n",
       "      <td>7.126822</td>\n",
       "      <td>6.822865</td>\n",
       "      <td>6.331795</td>\n",
       "      <td>5.795903</td>\n",
       "      <td>5.971143</td>\n",
       "      <td>5.381873</td>\n",
       "      <td>...</td>\n",
       "      <td>5.203703</td>\n",
       "      <td>6.778073</td>\n",
       "      <td>5.492302</td>\n",
       "      <td>5.322732</td>\n",
       "      <td>5.172698</td>\n",
       "      <td>5.255691</td>\n",
       "      <td>5.857087</td>\n",
       "      <td>5.282756</td>\n",
       "      <td>5.717348</td>\n",
       "      <td>5.695918</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ENSG00000259943</th>\n",
       "      <td>5.880182</td>\n",
       "      <td>5.451081</td>\n",
       "      <td>4.562843</td>\n",
       "      <td>5.745362</td>\n",
       "      <td>6.509814</td>\n",
       "      <td>5.059006</td>\n",
       "      <td>4.694381</td>\n",
       "      <td>5.897264</td>\n",
       "      <td>4.031779</td>\n",
       "      <td>6.163567</td>\n",
       "      <td>...</td>\n",
       "      <td>4.795975</td>\n",
       "      <td>5.629929</td>\n",
       "      <td>5.965515</td>\n",
       "      <td>5.146646</td>\n",
       "      <td>6.012784</td>\n",
       "      <td>5.250028</td>\n",
       "      <td>4.896769</td>\n",
       "      <td>5.316010</td>\n",
       "      <td>5.028731</td>\n",
       "      <td>4.513668</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ENSG00000250317</th>\n",
       "      <td>6.498122</td>\n",
       "      <td>7.056581</td>\n",
       "      <td>5.671896</td>\n",
       "      <td>6.774510</td>\n",
       "      <td>5.808357</td>\n",
       "      <td>6.491796</td>\n",
       "      <td>6.238326</td>\n",
       "      <td>5.942010</td>\n",
       "      <td>5.585238</td>\n",
       "      <td>6.953208</td>\n",
       "      <td>...</td>\n",
       "      <td>5.784468</td>\n",
       "      <td>6.190037</td>\n",
       "      <td>6.368464</td>\n",
       "      <td>5.839301</td>\n",
       "      <td>6.202581</td>\n",
       "      <td>5.832911</td>\n",
       "      <td>6.131257</td>\n",
       "      <td>6.585041</td>\n",
       "      <td>6.322645</td>\n",
       "      <td>5.890142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ENSG00000271852</th>\n",
       "      <td>1.409863</td>\n",
       "      <td>0.788887</td>\n",
       "      <td>1.544358</td>\n",
       "      <td>0.302041</td>\n",
       "      <td>0.185501</td>\n",
       "      <td>-1.533412</td>\n",
       "      <td>1.965067</td>\n",
       "      <td>-1.013507</td>\n",
       "      <td>0.520583</td>\n",
       "      <td>2.087955</td>\n",
       "      <td>...</td>\n",
       "      <td>2.287390</td>\n",
       "      <td>-0.862427</td>\n",
       "      <td>1.587465</td>\n",
       "      <td>1.599082</td>\n",
       "      <td>2.012894</td>\n",
       "      <td>1.301155</td>\n",
       "      <td>1.336574</td>\n",
       "      <td>1.762720</td>\n",
       "      <td>1.942644</td>\n",
       "      <td>1.481919</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4 rows × 297 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                 17-00963  18-00208  17-00256  18-00223  17-00908  17-00838  \\\n",
       "ENSG00000116918  5.527979  5.887300  5.654664  6.607494  7.126822  6.822865   \n",
       "ENSG00000259943  5.880182  5.451081  4.562843  5.745362  6.509814  5.059006   \n",
       "ENSG00000250317  6.498122  7.056581  5.671896  6.774510  5.808357  6.491796   \n",
       "ENSG00000271852  1.409863  0.788887  1.544358  0.302041  0.185501 -1.533412   \n",
       "\n",
       "                 18-00238  17-00011  17-00048  19-00025  ...  00-00015  \\\n",
       "ENSG00000116918  6.331795  5.795903  5.971143  5.381873  ...  5.203703   \n",
       "ENSG00000259943  4.694381  5.897264  4.031779  6.163567  ...  4.795975   \n",
       "ENSG00000250317  6.238326  5.942010  5.585238  6.953208  ...  5.784468   \n",
       "ENSG00000271852  1.965067 -1.013507  0.520583  2.087955  ...  2.287390   \n",
       "\n",
       "                 17-01023  17-01117  00-00004  17-00326  00-00005  17-00368  \\\n",
       "ENSG00000116918  6.778073  5.492302  5.322732  5.172698  5.255691  5.857087   \n",
       "ENSG00000259943  5.629929  5.965515  5.146646  6.012784  5.250028  4.896769   \n",
       "ENSG00000250317  6.190037  6.368464  5.839301  6.202581  5.832911  6.131257   \n",
       "ENSG00000271852 -0.862427  1.587465  1.599082  2.012894  1.301155  1.336574   \n",
       "\n",
       "                 19-00092  13-00126  17-00857  \n",
       "ENSG00000116918  5.282756  5.717348  5.695918  \n",
       "ENSG00000259943  5.316010  5.028731  4.513668  \n",
       "ENSG00000250317  6.585041  6.322645  5.890142  \n",
       "ENSG00000271852  1.762720  1.942644  1.481919  \n",
       "\n",
       "[4 rows x 297 columns]"
      ]
     },
     "execution_count": 227,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "BeatAML_w34_data_set3_df = BeatAML_w34_data[featureset_3]\n",
    "BeatAML_w34_data_set3_df_t = BeatAML_w34_data_set3_df.transpose()\n",
    "print(BeatAML_w34_data_set3_df_t.shape)\n",
    "BeatAML_w34_data_set3_df_t.head(4)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Processing features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "metadata": {},
   "outputs": [],
   "source": [
    "Data_dir = \"data/\"\n",
    "GE = pd.read_csv(Data_dir + \"RPKM.csv.gz\",index_col = \"Unnamed: 0\")\n",
    "\n",
    "#Formating gene expression data for training data\n",
    "GE.index = GE['Gene']\n",
    "\n",
    "# Define gene_map\n",
    "gene_map = GE.loc[:,['Gene','Symbol']]\n",
    "dic_gene_map = {}\n",
    "for i in range(0,gene_map.shape[0]):\n",
    "    dic_gene_map[gene_map.iloc[i,0]] = gene_map.iloc[i,1]\n",
    "    dic_gene_map[gene_map.iloc[i,1]] = gene_map.iloc[i,0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 229,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Load features from features that predict drug response, regulatory graph and house keeping genes\n",
    "#sele_house_keeping =  [\"GPS2\", \"RPS10\",\"ZNF91\",\"FNTA\",\"PSMC1\",\"GPS1\",\"MLH1\",\"ARAF\",\"SF3B2\",\"PSMD2\",\"SRP14\",\"GNB1\",\"HNRNPK\",\"ARF1\",\"RPS11\"]\n",
    "sele_house_keeping = ['RPS10','FNTA','COX6A1','BECN1', 'SF3B2','PSMB2','AUP1','SRP14','HNRNPK', 'CCNI', 'RHOA', 'PABPC1','RPS11','TPT1','FTL']\n",
    "#KG_expr_resp = pd.read_csv(\"./data/Supplementary_table_1_KG_Expr_drug_lasso_Coef_features_over0_gt50p.csv\")\n",
    "KG_expr_resp = pd.read_csv(\"data/Supplementary_table_1_KG_Expr_drug_lasso_Coef_features_over0_gt50p.csv\")\n",
    "#KG_reg = pd.read_csv(\"data/KG_regulatory_graph_Mar1_2023.csv\")\n",
    "KG_reg = pd.read_csv(\"data/KG_regulatory_graph_Mar11_2024.csv\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 230,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "44\n",
      "['CLASP1', 'HDAC5', 'CLIP4', 'NPC2', 'ACOT9', 'UBE2M', 'LILRB2', 'DPF2', 'ECHDC3', 'GDF11', 'MYO7A', 'FGD4', 'PAM', 'VENTX', 'ADPRHL1', 'SPICE1', 'NIPBL', 'SUN1', 'RFNG', 'CD14', 'LILRA3', 'MTSS1', 'BCL2', 'C1QA', 'GOLGA8J', 'LINC00471', 'PLA2G6', 'ZNF502', 'IGF2R', 'ZNF517', 'CASC10', 'RP1-257A7.4', 'AC138623.1', 'CNOT7P1', 'RP11-108L7.4', 'CDK2AP2P2', 'PDXP', 'EEF1GP2', 'CTD-3203P2.1', 'RP11-83N9.5', 'SMC5-AS1', 'ZSCAN16-AS1', 'RP11-355B11.2', 'CASC15']\n"
     ]
    }
   ],
   "source": [
    "# Feature selection: Selection of first neighbors\n",
    "#Drug = \"JQ1\"\n",
    "#Drug = \"Trametinib (GSK1120212)\"\n",
    "#Drug = \"Selumetinib (AZD6244)\"\n",
    "Drug = \"Venetoclax\"\n",
    "#Drug = \"Vandetanib (ZD6474)\"\n",
    "#Drug = \"Idelalisib\"\n",
    "\n",
    "\n",
    "res1 = KG_expr_resp.loc[KG_expr_resp['Drug'] == Drug]\n",
    "res1 = res1.sort_values(by = ['coef'])\n",
    "predictors = list(res1['Gene'])\n",
    "predictors.remove(Drug+\"_RSS\")\n",
    "predictors.remove(Drug+\"_intercept\")\n",
    "\n",
    "\n",
    "# Selection of the second neighbors\n",
    "res2 = KG_reg.loc[KG_reg['Target_gene'].isin( predictors)]\n",
    "res2 = res2.loc[res2['Coef'] <0]\n",
    "\n",
    "selected_genes = list(KG_expr_resp.loc[KG_expr_resp['Drug'] == Drug]['Symbol'])\n",
    "selected_genes.remove(Drug+\"_RSS\")\n",
    "selected_genes.remove(Drug+\"_intercept\")\n",
    "\n",
    "\n",
    "\n",
    "print(len(selected_genes))\n",
    "print(selected_genes)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Formating house keeping gene ids\n",
    "sele_house_keeping_id = []\n",
    "selected_genes_id = []\n",
    "\n",
    "for g in selected_genes:\n",
    "    selected_genes_id.append(dic_gene_map[g])\n",
    "    \n",
    "for g in sele_house_keeping:\n",
    "    if g in dic_gene_map:\n",
    "        sele_house_keeping_id.append(dic_gene_map[g])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "metadata": {},
   "outputs": [],
   "source": [
    "# drop columns with missing values\n",
    "training_data_beatAML_wave1_2 = BeatAML_w12_data_set3_df_t.dropna(axis=1)\n",
    "test_data_beatAML_wave3_4 = BeatAML_w34_data_set3_df_t.dropna(axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 233,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "House keeping genes:\n",
      "Regulators:\n",
      "House keeping genes:\n",
      "Regulators:\n"
     ]
    }
   ],
   "source": [
    "new_feature_df_t = formating_relative_expression(training_data_beatAML_wave1_2,selected_genes_id, sele_house_keeping_id, KG_reg, dic_gene_map= dic_gene_map)\n",
    "\n",
    "new_feature_df_test = formating_relative_expression(test_data_beatAML_wave3_4,selected_genes_id, sele_house_keeping_id, KG_reg, dic_gene_map= dic_gene_map)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "451\n",
      "(451, 273)\n",
      "(451, 232)\n",
      "(241, 232)\n",
      "(451, 232)\n"
     ]
    }
   ],
   "source": [
    "# New added function to filter features \n",
    "# features for the relative gene expression data in Beat AML wave 1 and 2\n",
    "print(new_feature_df_t.shape[0])\n",
    "\n",
    "new_feature_df_t_filter_low = new_feature_df_t.loc[:,new_feature_df_t.sum(axis = 0) > 0.05*new_feature_df_t.shape[0]]\n",
    "print(new_feature_df_t_filter_low.shape) # filter features with with more than 5% of the samples have non-zero values\n",
    "new_feature_df_test_filter_high = new_feature_df_t_filter_low.loc[:,new_feature_df_t_filter_low.sum(axis = 0) < 0.95*new_feature_df_t.shape[0]]\n",
    "print(new_feature_df_test_filter_high.shape) # filter features with with less than 95% of the samples have 1 values\n",
    "\n",
    "new_feature_df_t = new_feature_df_test_filter_high  # Relative gene expression feature for wave 1 and 2\n",
    "\n",
    "# features for the relative gene expression data in Beat AML wave 3 and 4\n",
    "\n",
    "new_feature_df_test = new_feature_df_test.loc[:,new_feature_df_t.columns] # Relative gene expression feature for wave 3 and 4\n",
    "\n",
    "print(new_feature_df_test.shape)\n",
    "print(new_feature_df_t.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 235,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(672, 17959)\n",
      "(297, 17959)\n"
     ]
    }
   ],
   "source": [
    "w12_f123 = BeatAML_w12_data.loc[:,featureset_1 + featureset_2 + featureset_3]\n",
    "w12_drug = BeatAML_w12_data.loc[:,[Drug, Drug+'_auc']]\n",
    "df_sele_w12 = pd.concat([ w12_f123,new_feature_df_t,w12_drug],axis = 1)\n",
    "\n",
    "w34_f123 = BeatAML_w34_data.loc[:,featureset_1 + featureset_2 + featureset_3]\n",
    "w34_drug = BeatAML_w34_data.loc[:,[Drug, Drug+'_auc']]\n",
    "df_sele_w34 = pd.concat([ w34_f123,new_feature_df_test,w34_drug],axis = 1)\n",
    "\n",
    "print(df_sele_w12.shape)\n",
    "print(df_sele_w34.shape)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 236,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "232\n",
      "232\n"
     ]
    }
   ],
   "source": [
    "BeatAML_w12_data_set4 = list(new_feature_df_t.columns)\n",
    "BeatAML_w34_data_set4 = list(new_feature_df_test.columns)\n",
    "print(len(BeatAML_w34_data_set4))\n",
    "print(len(BeatAML_w12_data_set4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {},
   "outputs": [],
   "source": [
    "#set1 = list(df_sele_w34.columns[0:23])\n",
    "#set2 = list(df_sele_w34.columns[23:40])\n",
    "#set3 = list(df_sele_w34.columns[41:41+19853]) \n",
    "set4 = BeatAML_w34_data_set4\n",
    "set1 = featureset_1 \n",
    "set2 = featureset_2 \n",
    "set3 = featureset_3\n",
    "\n",
    "dic_dataset_test = {\n",
    "    \"Mut\": df_sele_w34[set1 + [Drug, Drug+'_auc']].dropna(),\n",
    "    \"VAF\": df_sele_w34[set2 + [Drug, Drug+'_auc']].dropna(),\n",
    "    \"Expr\": pd.concat([df_sele_w34[set3], df_sele_w34[[Drug, Drug+'_auc']]], axis = 1).dropna(),\n",
    "    \"RelativeExpr\": df_sele_w34[set4 + [Drug, Drug+'_auc']].dropna(),\n",
    "    \"Mut+VAF\": pd.concat([df_sele_w34[set1], df_sele_w34[set2], df_sele_w34[[Drug, Drug+'_auc']]], axis = 1).dropna(),\n",
    "    \"Mut+RelativeExpr\": df_sele_w34[set1+ set4 + [Drug, Drug+'_auc']].dropna(),\n",
    "    \"Mut+VAF+RelativeExpr\": pd.concat([df_sele_w34[set1], df_sele_w34[set2], df_sele_w34[set4], df_sele_w34[[Drug, Drug+'_auc']]], axis = 1).dropna(),\n",
    "}\n",
    "\n",
    "dic_dataset_train = {\n",
    "    \"Mut\": df_sele_w12[set1 + [Drug, Drug+'_auc']].dropna(),\n",
    "    \"VAF\": df_sele_w12[set2 + [Drug, Drug+'_auc']].dropna(),\n",
    "    \"Expr\": pd.concat([df_sele_w12[set3], df_sele_w12[[Drug, Drug+'_auc']]], axis = 1).dropna(),\n",
    "    \"RelativeExpr\": df_sele_w12[set4 + [Drug, Drug+'_auc']].dropna(),\n",
    "    \"Mut+VAF\": pd.concat([df_sele_w12[set1], df_sele_w12[set2], df_sele_w12[[Drug, Drug+'_auc']]], axis = 1).dropna(),\n",
    "    \"Mut+RelativeExpr\": df_sele_w12[set1+ set4 + [Drug, Drug+'_auc']].dropna(),\n",
    "    \"Mut+VAF+RelativeExpr\": pd.concat([df_sele_w12[set1], df_sele_w12[set2], df_sele_w12[set4], df_sele_w12[[Drug, Drug+'_auc']]], axis = 1).dropna()\n",
    "    }\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Formatting training/testing labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get drugs - which drugs are we using? everything in the thresholds table.\n",
    "thresholds = pd.read_csv('data/selected_threshold_Apr22_2024.csv', index_col=0)\n",
    "all_drugs = thresholds.Drug"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Idelalisib', 'JQ1', 'Pazopanib (GW786034)', 'GDC-0941', '17-AAG (Tanespimycin)', 'Elesclomol', 'Venetoclax', 'Vandetanib (ZD6474)', 'Ponatinib (AP24534)', 'Vatalanib (PTK787)', 'Rapamycin', 'Dasatinib', 'Midostaurin', 'Neratinib (HKI-272)', 'Quizartinib (AC220)', 'Bortezomib (Velcade)', 'Flavopiridol', 'Cediranib (AZD2171)', 'Bosutinib (SKI-606)', 'Crenolanib', 'CYT387', 'KI20227', 'Vargetef', 'Tivozanib (AV-951)', 'Selinexor', 'Linifanib (ABT-869)', 'RAF265 (CHIR-265)', 'Tofacitinib (CP-690550)', 'Foretinib (XL880)', 'PP242', 'AZD1480', 'Cabozantinib', 'PI-103', 'BEZ235', 'Panobinostat', 'INK-128', 'JNJ-28312141', 'PD173955', 'JAK Inhibitor I', 'Axitinib (AG-013736)', 'CI-1040 (PD184352)', 'Doramapimod (BIRB 796)', 'VX-745', 'Trametinib (GSK1120212)', 'Masitinib (AB-1010)', 'NVP-TAE684', 'KW-2449', 'A-674563', 'Pelitinib (EKB-569)', 'PRT062607', 'Sorafenib', 'SNS-032 (BMS-387032)', 'GSK-1838705A', 'AT7519', 'MGCD-265', 'Dovitinib (CHIR-258)', 'MK-2206', 'Selumetinib (AZD6244)', 'Sunitinib', 'Lestaurtinib (CEP-701)', 'Ibrutinib (PCI-32765)', 'Motesanib (AMG-706)', 'NF-kB Activation Inhibitor']\n"
     ]
    }
   ],
   "source": [
    "# TODO: just use raw ic50 and AUC?\n",
    "print(all_drugs.to_list())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Running tests\n",
    "\n",
    "Axes along which we're doing the analysis:\n",
    "\n",
    "1. drugs (currently just venetoclax)\n",
    "2. methods (XGBoost regression, LASSOCV, kernel ridge regression (KRR))\n",
    "3. targets (AUC, IC50)\n",
    "4. feature sets (Mut, Expr, RelativeExpr, Mut+RelativeExpr)\n",
    "5. Cross-validation folds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import Lasso, LassoCV\n",
    "from sklearn.kernel_ridge import KernelRidge"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 286,
   "metadata": {},
   "outputs": [],
   "source": [
    "def run_regression_dataset(dataset, dic_dataset_train, dic_dataset_test, drug='Venetoclax', method='XGBoost', target='auc',\n",
    "                          model_kwargs=None):\n",
    "    \"\"\"\n",
    "    Runs 10-fold CV for a given drug, method, target, and feature set combo.\n",
    "    \"\"\"\n",
    "    print(dataset)\n",
    "    if method == 'XGBoost':\n",
    "        model = xgb.XGBRegressor\n",
    "    elif method == 'LassoCV':\n",
    "        model = LassoCV\n",
    "        if model_kwargs is None:\n",
    "            model_kwargs = {'selection': 'random', 'tol': 1e-3, 'n_jobs': 2}\n",
    "    elif method == 'Lasso':\n",
    "        model = Lasso\n",
    "    elif method == 'KernelRidge':\n",
    "        model = KernelRidge\n",
    "        if model_kwargs is None:\n",
    "            model_kwargs = {'kernel': 'rbf', }\n",
    "            if dataset == 'Expr':\n",
    "                model_kwargs['alpha'] = 0.1\n",
    "                model_kwargs['gamma'] = 1e-5\n",
    "    data_cur_train = dic_dataset_train[dataset]\n",
    "    data_cur_train = data_cur_train.dropna()\n",
    "\n",
    "    data_cur_test = dic_dataset_test[dataset][data_cur_train.columns]\n",
    "    data_cur_test = data_cur_test.dropna()\n",
    "\n",
    "    # TODO: get AUC as well?\n",
    "    result_all = {}\n",
    "\n",
    "    X = data_cur_train.iloc[:, :-2]\n",
    "    if target == 'ic50':\n",
    "        y = data_cur_train.loc[:, drug]\n",
    "        y_indep = data_cur_test.loc[:, drug]\n",
    "    elif target == 'auc':\n",
    "        y = data_cur_train.loc[:, drug+'_auc']\n",
    "        y_indep = data_cur_test.loc[:, drug+'_auc']\n",
    "    X_indep = data_cur_test.iloc[:,:-2]\n",
    "        \n",
    "    kfold = KFold(n_splits=10)\n",
    "    splits = kfold.split(X, y)\n",
    "\n",
    "    for i, (train_indices, test_indices) in enumerate(splits):\n",
    "        print('fold:', i)\n",
    "\n",
    "        train_data = X.iloc[train_indices]\n",
    "        test_data = X.iloc[test_indices]\n",
    "\n",
    "        train_labels = y.iloc[train_indices]\n",
    "        test_labels = y.iloc[test_indices]\n",
    "        t = time.time()\n",
    "        result_all[i] = model_run_with_Independent_testSet(train_data,\n",
    "                                                              test_data, \n",
    "                                                              train_labels, \n",
    "                                                              test_labels, \n",
    "                                                              y_indep,  \n",
    "                                                              X_indep,\n",
    "                                                          model=model,\n",
    "                                                          params=model_kwargs)\n",
    "        print('train_r2:', result_all[i]['train_r2'], 'test_r2:', result_all[i]['test_r2'],\n",
    "              'independent_r2:', result_all[i]['independent_r2'], 'runtime:', time.time() - t)\n",
    "        result_all[i]['runtime'] = time.time() - t\n",
    "    return result_all"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hyperparameter tuning for kernel ridge regression on gene expression data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 266,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.001 1e-06\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9973868711932986 test_r2: 0.3199069609674573 independent_r2: 0.9403130768460789 runtime: 2.535348892211914\n",
      "fold: 1\n",
      "train_r2: 0.9972062311709261 test_r2: 0.6801617787873158 independent_r2: 0.9639688914610457 runtime: 1.9499623775482178\n",
      "fold: 2\n",
      "train_r2: 0.9972092291215474 test_r2: 0.7416748963768485 independent_r2: 0.975384049812695 runtime: 2.8780786991119385\n",
      "fold: 3\n",
      "train_r2: 0.9975016621521886 test_r2: 0.3625562326463203 independent_r2: 0.9600470735742347 runtime: 2.692117214202881\n",
      "fold: 4\n",
      "train_r2: 0.9975469823109459 test_r2: 0.5763439271078258 independent_r2: 0.9577353761539086 runtime: 1.8797328472137451\n",
      "fold: 5\n",
      "train_r2: 0.9973275142640408 test_r2: 0.49400926661426703 independent_r2: 0.9631815097728874 runtime: 2.3505237102508545\n",
      "fold: 6\n",
      "train_r2: 0.997635894180863 test_r2: 0.4165620679682732 independent_r2: 0.9470636592718319 runtime: 3.083024501800537\n",
      "fold: 7\n",
      "train_r2: 0.9972919453435617 test_r2: 0.4931844510313057 independent_r2: 0.9363315243465566 runtime: 2.5739834308624268\n",
      "fold: 8\n",
      "train_r2: 0.9974875342579299 test_r2: 0.5327792959712436 independent_r2: 0.9464481506885447 runtime: 2.0572409629821777\n",
      "fold: 9\n",
      "train_r2: 0.9976303594147318 test_r2: 0.37361803163125573 independent_r2: 0.8957090117115207 runtime: 1.8231971263885498\n",
      "0.001 1e-05\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9999738636307893 test_r2: 0.37452171905448406 independent_r2: 0.9472714159517305 runtime: 1.3175606727600098\n",
      "fold: 1\n",
      "train_r2: 0.9999727832469159 test_r2: 0.6677713504202489 independent_r2: 0.96513429768155 runtime: 1.6808900833129883\n",
      "fold: 2\n",
      "train_r2: 0.9999727019912643 test_r2: 0.7522094564700701 independent_r2: 0.9788111730077911 runtime: 2.6602413654327393\n",
      "fold: 3\n",
      "train_r2: 0.9999754062878524 test_r2: 0.36948835883820597 independent_r2: 0.962766121436687 runtime: 2.2631618976593018\n",
      "fold: 4\n",
      "train_r2: 0.9999763450743798 test_r2: 0.5529848869466317 independent_r2: 0.9577207475060521 runtime: 1.588155746459961\n",
      "fold: 5\n",
      "train_r2: 0.99997360733 test_r2: 0.5267929152298818 independent_r2: 0.9678713268661535 runtime: 2.4681406021118164\n",
      "fold: 6\n",
      "train_r2: 0.9999762330149052 test_r2: 0.45314788895393676 independent_r2: 0.9523745688479226 runtime: 3.1751410961151123\n",
      "fold: 7\n",
      "train_r2: 0.9999734240763437 test_r2: 0.5304031526100347 independent_r2: 0.9431891059209567 runtime: 1.3301913738250732\n",
      "fold: 8\n",
      "train_r2: 0.9999746389571771 test_r2: 0.5695462004005318 independent_r2: 0.9527002380710391 runtime: 3.131037950515747\n",
      "fold: 9\n",
      "train_r2: 0.9999759006739652 test_r2: 0.363687244009237 independent_r2: 0.8960454952644717 runtime: 3.118426561355591\n",
      "0.001 0.0001\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9999987505457006 test_r2: -0.6097015965821142 independent_r2: 0.8643603145764068 runtime: 2.0585379600524902\n",
      "fold: 1\n",
      "train_r2: 0.9999987110893583 test_r2: -0.11814053654411572 independent_r2: 0.8827373147159633 runtime: 2.48874568939209\n",
      "fold: 2\n",
      "train_r2: 0.9999987041540981 test_r2: 0.15682470940421367 independent_r2: 0.9279829717508339 runtime: 2.302713394165039\n",
      "fold: 3\n",
      "train_r2: 0.9999987515182316 test_r2: -0.16093328738678458 independent_r2: 0.931484019817613 runtime: 2.521314859390259\n",
      "fold: 4\n",
      "train_r2: 0.9999987588040267 test_r2: -0.21871663777569972 independent_r2: 0.8847893567337047 runtime: 2.9434540271759033\n",
      "fold: 5\n",
      "train_r2: 0.9999987265425684 test_r2: 0.11496448275994897 independent_r2: 0.93995480278959 runtime: 2.7563815116882324\n",
      "fold: 6\n",
      "train_r2: 0.9999987547945415 test_r2: -0.2268716428318247 independent_r2: 0.8931988763592623 runtime: 2.696803331375122\n",
      "fold: 7\n",
      "train_r2: 0.9999987432759531 test_r2: -0.2960538260618195 independent_r2: 0.8432693435094443 runtime: 2.0738699436187744\n",
      "fold: 8\n",
      "train_r2: 0.9999986989729879 test_r2: 0.3683466475939231 independent_r2: 0.9306237003844122 runtime: 2.7589495182037354\n",
      "fold: 9\n",
      "train_r2: 0.9999987612510152 test_r2: -0.28867550820573595 independent_r2: 0.7895086202143885 runtime: 2.695021629333496\n",
      "0.001 0.001\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9999944534214531 test_r2: -6.912778200920371 independent_r2: 0.3332392317885877 runtime: 2.2087838649749756\n",
      "fold: 1\n",
      "train_r2: 0.9999943439734454 test_r2: -5.481526123209863 independent_r2: 0.32026491785091193 runtime: 1.6349313259124756\n",
      "fold: 2\n",
      "train_r2: 0.9999943636936205 test_r2: -5.602218271335537 independent_r2: 0.4360974913221283 runtime: 1.9101386070251465\n",
      "fold: 3\n",
      "train_r2: 0.9999943464670279 test_r2: -6.440641840778795 independent_r2: 0.5608702680529917 runtime: 2.7915515899658203\n",
      "fold: 4\n",
      "train_r2: 0.9999942385852435 test_r2: -4.46046955304361 independent_r2: 0.4837976190868697 runtime: 1.9832022190093994\n",
      "fold: 5\n",
      "train_r2: 0.9999944557039838 test_r2: -7.140497293530364 independent_r2: 0.4477140538686495 runtime: 2.746263027191162\n",
      "fold: 6\n",
      "train_r2: 0.9999942170721933 test_r2: -4.339593670607452 independent_r2: 0.5351795663993546 runtime: 2.5496556758880615\n",
      "fold: 7\n",
      "train_r2: 0.9999941732728553 test_r2: -3.940205106752485 independent_r2: 0.40258440302829024 runtime: 2.6586451530456543\n",
      "fold: 8\n",
      "train_r2: 0.9999942177454874 test_r2: -4.202538831711741 independent_r2: 0.42859462808354043 runtime: 2.1566519737243652\n",
      "fold: 9\n",
      "train_r2: 0.9999939623749983 test_r2: -3.1462904444561417 independent_r2: 0.32274602943717745 runtime: 1.8928818702697754\n",
      "0.001 0.01\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9999944531700704 test_r2: -6.9131067166781435 independent_r2: 0.33321154978290457 runtime: 2.2073230743408203\n",
      "fold: 1\n",
      "train_r2: 0.9999943438865703 test_r2: -5.4825999932237846 independent_r2: 0.3201522989709541 runtime: 1.3276968002319336\n",
      "fold: 2\n",
      "train_r2: 0.9999943636053702 test_r2: -5.603517167631292 independent_r2: 0.43598655213669457 runtime: 1.3822462558746338\n",
      "fold: 3\n",
      "train_r2: 0.9999943461470522 test_r2: -6.440645858041065 independent_r2: 0.5608700306673869 runtime: 2.8128905296325684\n",
      "fold: 4\n",
      "train_r2: 0.9999942382624072 test_r2: -4.4605108592166545 independent_r2: 0.48379371397916293 runtime: 2.3806095123291016\n",
      "fold: 5\n",
      "train_r2: 0.9999944553845824 test_r2: -7.140509497533987 independent_r2: 0.44771322560728166 runtime: 1.329718828201294\n",
      "fold: 6\n",
      "train_r2: 0.9999942167474036 test_r2: -4.3396179246291755 independent_r2: 0.5351774547755773 runtime: 1.4006924629211426\n",
      "fold: 7\n",
      "train_r2: 0.9999941729329251 test_r2: -3.9402069603018353 independent_r2: 0.40258417858299744 runtime: 3.420269727706909\n",
      "fold: 8\n",
      "train_r2: 0.9999942174720101 test_r2: -4.202792103195392 independent_r2: 0.4285668107709032 runtime: 2.2672550678253174\n",
      "fold: 9\n",
      "train_r2: 0.9999939620182945 test_r2: -3.146294814083329 independent_r2: 0.322745315410903 runtime: 1.4862060546875\n",
      "0.001 0.1\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9999944531688525 test_r2: -6.9131067166781435 independent_r2: 0.3332115497817959 runtime: 1.997948408126831\n",
      "fold: 1\n",
      "train_r2: 0.9999943438853613 test_r2: -5.4825999932237846 independent_r2: 0.32015229896987774 runtime: 2.443157911300659\n",
      "fold: 2\n",
      "train_r2: 0.9999943636041063 test_r2: -5.603517167631292 independent_r2: 0.4359865521355394 runtime: 1.448329210281372\n",
      "fold: 3\n",
      "train_r2: 0.9999943461457661 test_r2: -6.440645858041065 independent_r2: 0.5608700306661867 runtime: 2.3441274166107178\n",
      "fold: 4\n",
      "train_r2: 0.9999942382610515 test_r2: -4.4605108592166545 independent_r2: 0.48379371397793947 runtime: 3.007716178894043\n",
      "fold: 5\n",
      "train_r2: 0.9999944553833702 test_r2: -7.140509497533987 independent_r2: 0.44771322560615445 runtime: 2.6463279724121094\n",
      "fold: 6\n",
      "train_r2: 0.9999942167461628 test_r2: -4.3396179246291755 independent_r2: 0.5351774547744509 runtime: 2.2641613483428955\n",
      "fold: 7\n",
      "train_r2: 0.9999941729315484 test_r2: -3.9402069603018353 independent_r2: 0.40258417858178563 runtime: 1.6280748844146729\n",
      "fold: 8\n",
      "train_r2: 0.9999942174707105 test_r2: -4.202792103195392 independent_r2: 0.42856681076974457 runtime: 2.3527772426605225\n",
      "fold: 9\n",
      "train_r2: 0.9999939620168466 test_r2: -3.146294814083329 independent_r2: 0.3227453154096922 runtime: 1.6249470710754395\n",
      "0.001 1\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9999944531566732 test_r2: -6.9131067166781435 independent_r2: 0.3332115497707093 runtime: 1.559253215789795\n",
      "fold: 1\n",
      "train_r2: 0.9999943438732729 test_r2: -5.4825999932237846 independent_r2: 0.32015229895911446 runtime: 2.6646575927734375\n",
      "fold: 2\n",
      "train_r2: 0.9999943635914675 test_r2: -5.603517167631292 independent_r2: 0.43598655212398707 runtime: 1.9357993602752686\n",
      "fold: 3\n",
      "train_r2: 0.999994346132905 test_r2: -6.440645858041065 independent_r2: 0.5608700306541838 runtime: 1.8428645133972168\n",
      "fold: 4\n",
      "train_r2: 0.9999942382474939 test_r2: -4.4605108592166545 independent_r2: 0.48379371396570436 runtime: 2.7537503242492676\n",
      "fold: 5\n",
      "train_r2: 0.9999944553712475 test_r2: -7.140509497533987 independent_r2: 0.44771322559488214 runtime: 3.084595203399658\n",
      "fold: 6\n",
      "train_r2: 0.9999942167337554 test_r2: -4.3396179246291755 independent_r2: 0.5351774547631859 runtime: 2.8739407062530518\n",
      "fold: 7\n",
      "train_r2: 0.9999941729177795 test_r2: -3.9402069603018353 independent_r2: 0.40258417856966666 runtime: 1.5000519752502441\n",
      "fold: 8\n",
      "train_r2: 0.9999942174577147 test_r2: -4.202792103195392 independent_r2: 0.42856681075815883 runtime: 2.474034070968628\n",
      "fold: 9\n",
      "train_r2: 0.9999939620023682 test_r2: -3.146294814083329 independent_r2: 0.3227453153975852 runtime: 2.126946449279785\n",
      "0.01 1e-06\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.932317177771831 test_r2: 0.34771629297202533 independent_r2: 0.8833885341614379 runtime: 2.9434385299682617\n",
      "fold: 1\n",
      "train_r2: 0.9275359383356488 test_r2: 0.6886560981513704 independent_r2: 0.9027881035761443 runtime: 2.3400349617004395\n",
      "fold: 2\n",
      "train_r2: 0.9277619556585659 test_r2: 0.7604561713171529 independent_r2: 0.9134744239543668 runtime: 1.9815938472747803\n",
      "fold: 3\n",
      "train_r2: 0.9319766009487301 test_r2: 0.5099040898629188 independent_r2: 0.9075570676247229 runtime: 2.5154237747192383\n",
      "fold: 4\n",
      "train_r2: 0.9306681628016735 test_r2: 0.6372655521504751 independent_r2: 0.9031065667156714 runtime: 2.6893255710601807\n",
      "fold: 5\n",
      "train_r2: 0.9312055299330613 test_r2: 0.527520939820768 independent_r2: 0.9038419412797641 runtime: 2.6121840476989746\n",
      "fold: 6\n",
      "train_r2: 0.9339317962003905 test_r2: 0.5297384197364898 independent_r2: 0.8990435109433795 runtime: 2.3220126628875732\n",
      "fold: 7\n",
      "train_r2: 0.9319136174488302 test_r2: 0.4975445910205134 independent_r2: 0.8793968842613912 runtime: 2.6457021236419678\n",
      "fold: 8\n",
      "train_r2: 0.9329273651216548 test_r2: 0.5586117186456498 independent_r2: 0.8918159765933573 runtime: 1.5243275165557861\n",
      "fold: 9\n",
      "train_r2: 0.9355132676775545 test_r2: 0.3989338790883524 independent_r2: 0.8479753207026962 runtime: 2.9477219581604004\n",
      "0.01 1e-05\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9978616297781636 test_r2: 0.3792192872929917 independent_r2: 0.9457433484413194 runtime: 1.3229906558990479\n",
      "fold: 1\n",
      "train_r2: 0.9977683100535135 test_r2: 0.6702956897854646 independent_r2: 0.9634349907856772 runtime: 2.659576654434204\n",
      "fold: 2\n",
      "train_r2: 0.9977654615704813 test_r2: 0.7545610670436484 independent_r2: 0.976993369621922 runtime: 2.015573024749756\n",
      "fold: 3\n",
      "train_r2: 0.9979668771564566 test_r2: 0.3954480527566516 independent_r2: 0.9624226593175675 runtime: 2.566291093826294\n",
      "fold: 4\n",
      "train_r2: 0.9980198418588714 test_r2: 0.5684008369600475 independent_r2: 0.9574114547115521 runtime: 3.3251118659973145\n",
      "fold: 5\n",
      "train_r2: 0.9978471405095736 test_r2: 0.5301446852610174 independent_r2: 0.9661172601683617 runtime: 2.711221218109131\n",
      "fold: 6\n",
      "train_r2: 0.9980275890145389 test_r2: 0.47124636384243335 independent_r2: 0.9521798002251665 runtime: 1.9611804485321045\n",
      "fold: 7\n",
      "train_r2: 0.9978332436229675 test_r2: 0.5296425702001799 independent_r2: 0.9412160962059161 runtime: 3.0174124240875244\n",
      "fold: 8\n",
      "train_r2: 0.9979172570012359 test_r2: 0.5722732145226831 independent_r2: 0.9511683339373252 runtime: 2.089890718460083\n",
      "fold: 9\n",
      "train_r2: 0.9980146634369497 test_r2: 0.36666144134302436 independent_r2: 0.8948936324502071 runtime: 2.1269443035125732\n",
      "0.01 0.0001\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9998770823213391 test_r2: -0.6171413295840802 independent_r2: 0.8636225992979327 runtime: 2.0051302909851074\n",
      "fold: 1\n",
      "train_r2: 0.9998731836998388 test_r2: -0.1233867312810093 independent_r2: 0.8820752996719831 runtime: 3.268023729324341\n",
      "fold: 2\n",
      "train_r2: 0.9998725081480064 test_r2: 0.15086863765639624 independent_r2: 0.9273588480618127 runtime: 2.353039264678955\n",
      "fold: 3\n",
      "train_r2: 0.9998771598481347 test_r2: -0.16507909129194953 independent_r2: 0.9311258069040428 runtime: 2.5096383094787598\n",
      "fold: 4\n",
      "train_r2: 0.9998778405886242 test_r2: -0.22246942969466343 independent_r2: 0.8843254096784529 runtime: 1.5782294273376465\n",
      "fold: 5\n",
      "train_r2: 0.9998747046670694 test_r2: 0.1094502074265723 independent_r2: 0.9394651295911265 runtime: 2.9270172119140625\n",
      "fold: 6\n",
      "train_r2: 0.9998774673169059 test_r2: -0.23054838153224488 independent_r2: 0.8927686297495265 runtime: 2.3191041946411133\n",
      "fold: 7\n",
      "train_r2: 0.9998763360831264 test_r2: -0.3008262309282932 independent_r2: 0.8425846394753903 runtime: 1.7321937084197998\n",
      "fold: 8\n",
      "train_r2: 0.9998719482593919 test_r2: 0.36523984830276235 independent_r2: 0.9301696480142563 runtime: 2.9375522136688232\n",
      "fold: 9\n",
      "train_r2: 0.9998780791161083 test_r2: -0.29192172331528443 independent_r2: 0.788877616783253 runtime: 2.226604461669922\n",
      "0.01 0.001\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9994551828802355 test_r2: -6.9127811280490805 independent_r2: 0.3327477979040262 runtime: 2.3208606243133545\n",
      "fold: 1\n",
      "train_r2: 0.9994444324060586 test_r2: -5.481535625616499 independent_r2: 0.31977398663401513 runtime: 1.608550786972046\n",
      "fold: 2\n",
      "train_r2: 0.9994463694341439 test_r2: -5.6022297693255 independent_r2: 0.4355953353179939 runtime: 2.7805304527282715\n",
      "fold: 3\n",
      "train_r2: 0.9994446771402593 test_r2: -6.44064187657613 independent_r2: 0.5603569953419583 runtime: 2.731414794921875\n",
      "fold: 4\n",
      "train_r2: 0.9994340803674541 test_r2: -4.4604699211020735 independent_r2: 0.4832917918385703 runtime: 2.8489577770233154\n",
      "fold: 5\n",
      "train_r2: 0.9994554070239425 test_r2: -7.140497402273214 independent_r2: 0.4472117555781503 runtime: 2.9671108722686768\n",
      "fold: 6\n",
      "train_r2: 0.9994319672298557 test_r2: -4.3395938867286565 independent_r2: 0.5346687696230134 runtime: 2.117159605026245\n",
      "fold: 7\n",
      "train_r2: 0.9994276649935272 test_r2: -3.940205122791128 independent_r2: 0.4020864892336724 runtime: 2.393773078918457\n",
      "fold: 8\n",
      "train_r2: 0.9994320334097293 test_r2: -4.202541088202867 independent_r2: 0.4280939419263292 runtime: 1.2852976322174072\n",
      "fold: 9\n",
      "train_r2: 0.9994069493764159 test_r2: -3.1462904833910077 independent_r2: 0.3222558580348093 runtime: 1.2522687911987305\n",
      "0.01 0.01\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9994551584144502 test_r2: -6.9131067166781435 independent_r2: 0.33272034049210686 runtime: 3.5078089237213135\n",
      "fold: 1\n",
      "train_r2: 0.9994444239592565 test_r2: -5.4825999932237846 independent_r2: 0.31966235684147914 runtime: 2.576237678527832\n",
      "fold: 2\n",
      "train_r2: 0.9994463608539502 test_r2: -5.603517167631292 independent_r2: 0.43548537041278046 runtime: 2.6313817501068115\n",
      "fold: 3\n",
      "train_r2: 0.9994446459975759 test_r2: -6.440645858041065 independent_r2: 0.5603567312873567 runtime: 2.5832624435424805\n",
      "fold: 4\n",
      "train_r2: 0.9994340489467826 test_r2: -4.4605108592166545 independent_r2: 0.4832878934450543 runtime: 2.521692991256714\n",
      "fold: 5\n",
      "train_r2: 0.999455375936522 test_r2: -7.140509497533987 independent_r2: 0.447210906024347 runtime: 1.2913122177124023\n",
      "fold: 6\n",
      "train_r2: 0.9994319356180273 test_r2: -4.3396179246291755 independent_r2: 0.5346666483899045 runtime: 2.1059889793395996\n",
      "fold: 7\n",
      "train_r2: 0.9994276319088251 test_r2: -3.9402069603018353 independent_r2: 0.4020862379480771 runtime: 1.996948480606079\n",
      "fold: 8\n",
      "train_r2: 0.9994320067936292 test_r2: -4.202792103195392 independent_r2: 0.42806634899709506 runtime: 2.370471715927124\n",
      "fold: 9\n",
      "train_r2: 0.9994069146591703 test_r2: -3.146294814083329 independent_r2: 0.32225512167650605 runtime: 2.8850901126861572\n",
      "0.01 0.1\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.999455158402487 test_r2: -6.9131067166781435 independent_r2: 0.3327203404812169 runtime: 2.1609246730804443\n",
      "fold: 1\n",
      "train_r2: 0.9994444239473825 test_r2: -5.4825999932237846 independent_r2: 0.3196623568309067 runtime: 2.0775809288024902\n",
      "fold: 2\n",
      "train_r2: 0.9994463608415357 test_r2: -5.603517167631292 independent_r2: 0.4354853704014332 runtime: 2.328354835510254\n",
      "fold: 3\n",
      "train_r2: 0.9994446459849432 test_r2: -6.440645858041065 independent_r2: 0.5603567312755671 runtime: 2.576545000076294\n",
      "fold: 4\n",
      "train_r2: 0.9994340489334655 test_r2: -4.4605108592166545 independent_r2: 0.48328789343303624 runtime: 1.366248369216919\n",
      "fold: 5\n",
      "train_r2: 0.9994553759246144 test_r2: -7.140509497533987 independent_r2: 0.44721090601327484 runtime: 2.7888686656951904\n",
      "fold: 6\n",
      "train_r2: 0.9994319356058401 test_r2: -4.3396179246291755 independent_r2: 0.5346666483788394 runtime: 1.6745917797088623\n",
      "fold: 7\n",
      "train_r2: 0.9994276318953007 test_r2: -3.9402069603018353 independent_r2: 0.40208623793617326 runtime: 1.5279386043548584\n",
      "fold: 8\n",
      "train_r2: 0.999432006780864 test_r2: -4.202792103195392 independent_r2: 0.42806634898571483 runtime: 1.259275197982788\n",
      "fold: 9\n",
      "train_r2: 0.9994069146449488 test_r2: -3.146294814083329 independent_r2: 0.3222551216646138 runtime: 2.333137035369873\n",
      "0.01 1\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9994551582828548 test_r2: -6.9131067166781435 independent_r2: 0.33272034037231646 runtime: 1.3202590942382812\n",
      "fold: 1\n",
      "train_r2: 0.9994444238286431 test_r2: -5.4825999932237846 independent_r2: 0.31966235672518284 runtime: 1.8133068084716797\n",
      "fold: 2\n",
      "train_r2: 0.9994463607173907 test_r2: -5.603517167631292 independent_r2: 0.43548537028795986 runtime: 1.3043992519378662\n",
      "fold: 3\n",
      "train_r2: 0.9994446458586155 test_r2: -6.440645858041065 independent_r2: 0.5603567311576698 runtime: 1.449091911315918\n",
      "fold: 4\n",
      "train_r2: 0.9994340488002962 test_r2: -4.4605108592166545 independent_r2: 0.4832878933128565 runtime: 1.4484717845916748\n",
      "fold: 5\n",
      "train_r2: 0.9994553758055392 test_r2: -7.140509497533987 independent_r2: 0.44721090590255297 runtime: 2.443138360977173\n",
      "fold: 6\n",
      "train_r2: 0.9994319354839675 test_r2: -4.3396179246291755 independent_r2: 0.5346666482681888 runtime: 2.451355457305908\n",
      "fold: 7\n",
      "train_r2: 0.9994276317600558 test_r2: -3.9402069603018353 independent_r2: 0.40208623781713604 runtime: 2.487281084060669\n",
      "fold: 8\n",
      "train_r2: 0.999432006653212 test_r2: -4.202792103195392 independent_r2: 0.4280663488719144 runtime: 1.702500581741333\n",
      "fold: 9\n",
      "train_r2: 0.999406914502733 test_r2: -3.146294814083329 independent_r2: 0.322255121545693 runtime: 3.070803165435791\n",
      "0.1 1e-06\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.7097025921434981 test_r2: 0.32116982892224033 independent_r2: 0.6783861786010422 runtime: 2.3929860591888428\n",
      "fold: 1\n",
      "train_r2: 0.6858403468257039 test_r2: 0.6678843371809902 independent_r2: 0.6852750106897432 runtime: 3.336918830871582\n",
      "fold: 2\n",
      "train_r2: 0.6873481805549221 test_r2: 0.7277365741891995 independent_r2: 0.6908068679208302 runtime: 3.25813889503479\n",
      "fold: 3\n",
      "train_r2: 0.6953596410523988 test_r2: 0.5158976002574367 independent_r2: 0.6869624407086037 runtime: 2.952824831008911\n",
      "fold: 4\n",
      "train_r2: 0.6918903275752655 test_r2: 0.5992795142398031 independent_r2: 0.6839121946994031 runtime: 2.657087802886963\n",
      "fold: 5\n",
      "train_r2: 0.700131457923086 test_r2: 0.5317292327809847 independent_r2: 0.6888103321556571 runtime: 2.199033737182617\n",
      "fold: 6\n",
      "train_r2: 0.6930366276841534 test_r2: 0.6251037517668869 independent_r2: 0.6885030136905824 runtime: 1.9251396656036377\n",
      "fold: 7\n",
      "train_r2: 0.7037949230279457 test_r2: 0.46114447786088275 independent_r2: 0.6744984488550332 runtime: 2.1382946968078613\n",
      "fold: 8\n",
      "train_r2: 0.6999708983528508 test_r2: 0.572319751645699 independent_r2: 0.6859512270778441 runtime: 1.642843246459961\n",
      "fold: 9\n",
      "train_r2: 0.7149608770655211 test_r2: 0.35652102809932495 independent_r2: 0.6568829699073624 runtime: 2.7553138732910156\n",
      "0.1 1e-05\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.936442692727501 test_r2: 0.38930620188421217 independent_r2: 0.8906507005865556 runtime: 1.333787202835083\n",
      "fold: 1\n",
      "train_r2: 0.9329876308066478 test_r2: 0.678556577213314 independent_r2: 0.9065860433855352 runtime: 1.4606373310089111\n",
      "fold: 2\n",
      "train_r2: 0.9332941932511114 test_r2: 0.7576768104783544 independent_r2: 0.9182966044685126 runtime: 1.792074203491211\n",
      "fold: 3\n",
      "train_r2: 0.9367054786027903 test_r2: 0.49111968208496337 independent_r2: 0.9108642025925191 runtime: 1.0645687580108643\n",
      "fold: 4\n",
      "train_r2: 0.9361181009929956 test_r2: 0.6223674311825877 independent_r2: 0.9066191937828932 runtime: 2.213867664337158\n",
      "fold: 5\n",
      "train_r2: 0.936112095327022 test_r2: 0.5504831436086064 independent_r2: 0.9099717665950133 runtime: 1.1415910720825195\n",
      "fold: 6\n",
      "train_r2: 0.937489114108814 test_r2: 0.553774721558148 independent_r2: 0.9043675457019177 runtime: 1.9905669689178467\n",
      "fold: 7\n",
      "train_r2: 0.9362013480896815 test_r2: 0.5231593176292204 independent_r2: 0.8862629738881675 runtime: 1.3370981216430664\n",
      "fold: 8\n",
      "train_r2: 0.9367678477920388 test_r2: 0.5820607564950611 independent_r2: 0.8978100822300762 runtime: 1.4133610725402832\n",
      "fold: 9\n",
      "train_r2: 0.9393174732717441 test_r2: 0.3687958561885679 independent_r2: 0.8462291919816248 runtime: 1.5717799663543701\n",
      "0.1 0.0001\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9894523680844711 test_r2: -0.6902109194982091 independent_r2: 0.8479703154321416 runtime: 1.9143497943878174\n",
      "fold: 1\n",
      "train_r2: 0.9891050817645116 test_r2: -0.17531610230850547 independent_r2: 0.8670357035915222 runtime: 3.053365707397461\n",
      "fold: 2\n",
      "train_r2: 0.9890530291371814 test_r2: 0.09175364067775738 independent_r2: 0.9124147500431854 runtime: 3.6843807697296143\n",
      "fold: 3\n",
      "train_r2: 0.9894446024486825 test_r2: -0.206951663035982 independent_r2: 0.9189128922574816 runtime: 3.1051156520843506\n",
      "fold: 4\n",
      "train_r2: 0.989477983948912 test_r2: -0.26008668266939305 independent_r2: 0.8713788140793362 runtime: 1.1918492317199707\n",
      "fold: 5\n",
      "train_r2: 0.9892358446887878 test_r2: 0.054113584974744944 independent_r2: 0.9257974976008871 runtime: 1.4554064273834229\n",
      "fold: 6\n",
      "train_r2: 0.9894606403889372 test_r2: -0.26729099085272146 independent_r2: 0.8801069559623452 runtime: 1.4976401329040527\n",
      "fold: 7\n",
      "train_r2: 0.9893657242891003 test_r2: -0.3476102496018092 independent_r2: 0.8276892087626302 runtime: 1.355806827545166\n",
      "fold: 8\n",
      "train_r2: 0.9889653152328073 test_r2: 0.33333327707439075 independent_r2: 0.9169565840815882 runtime: 2.1523516178131104\n",
      "fold: 9\n",
      "train_r2: 0.9894912463923149 test_r2: -0.32385490988425647 independent_r2: 0.7749885295527679 runtime: 2.6271307468414307\n",
      "0.1 0.001\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9540685992120085 test_r2: -6.912807765118183 independent_r2: 0.2914058011519126 runtime: 2.0271213054656982\n",
      "fold: 1\n",
      "train_r2: 0.9531623795863893 test_r2: -5.48162215266186 independent_r2: 0.2785306773755538 runtime: 2.8865370750427246\n",
      "fold: 2\n",
      "train_r2: 0.9533256813404667 test_r2: -5.6023344641133415 independent_r2: 0.39340622923393986 runtime: 2.897024154663086\n",
      "fold: 3\n",
      "train_r2: 0.9531828584417752 test_r2: -6.440642202331923 independent_r2: 0.517158583539251 runtime: 3.2430903911590576\n",
      "fold: 4\n",
      "train_r2: 0.9522894873603114 test_r2: -4.460473270447602 independent_r2: 0.4407224676472523 runtime: 2.964031457901001\n",
      "fold: 5\n",
      "train_r2: 0.954087450096139 test_r2: -7.140498391837845 independent_r2: 0.404937375658615 runtime: 1.1940488815307617\n",
      "fold: 6\n",
      "train_r2: 0.9521113363110383 test_r2: -4.339595853434703 independent_r2: 0.4916799950601888 runtime: 1.455669641494751\n",
      "fold: 7\n",
      "train_r2: 0.951748622829628 test_r2: -3.940205269137885 independent_r2: 0.36018071093374704 runtime: 2.0500447750091553\n",
      "fold: 8\n",
      "train_r2: 0.9521169501648751 test_r2: -4.2025616225889895 independent_r2: 0.38597332885347313 runtime: 2.1516048908233643\n",
      "fold: 9\n",
      "train_r2: 0.9500021653817959 test_r2: -3.1462908377002696 independent_r2: 0.28100203215396424 runtime: 1.6385753154754639\n",
      "0.1 0.01\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9540667024628893 test_r2: -6.9131067166781435 independent_r2: 0.2913788829198817 runtime: 2.2158565521240234\n",
      "fold: 1\n",
      "train_r2: 0.9531617257891526 test_r2: -5.4825999932237846 independent_r2: 0.2784275468754137 runtime: 2.447550058364868\n",
      "fold: 2\n",
      "train_r2: 0.9533250172215686 test_r2: -5.603517167631292 independent_r2: 0.393304606810684 runtime: 3.054903507232666\n",
      "fold: 3\n",
      "train_r2: 0.9531804449103993 test_r2: -6.440645858041065 independent_r2: 0.5171561140812522 runtime: 1.6309502124786377\n",
      "fold: 4\n",
      "train_r2: 0.9522870522290313 test_r2: -4.4605108592166545 independent_r2: 0.4407167154588635 runtime: 3.225377321243286\n",
      "fold: 5\n",
      "train_r2: 0.9540850408314628 test_r2: -7.140509497533987 independent_r2: 0.40493437723128967 runtime: 2.321800708770752\n",
      "fold: 6\n",
      "train_r2: 0.9521088863795788 test_r2: -4.3396179246291755 independent_r2: 0.49167584809868037 runtime: 2.5692062377929688\n",
      "fold: 7\n",
      "train_r2: 0.9517460588076141 test_r2: -3.9402069603018353 independent_r2: 0.36017825287026883 runtime: 2.582411289215088\n",
      "fold: 8\n",
      "train_r2: 0.9521148868994465 test_r2: -4.202792103195392 independent_r2: 0.3859461782777053 runtime: 1.457259178161621\n",
      "fold: 9\n",
      "train_r2: 0.9499994748157329 test_r2: -3.146294814083329 independent_r2: 0.28099913599266113 runtime: 2.3440661430358887\n",
      "0.1 0.1\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9540667023620324 test_r2: -6.9131067166781435 independent_r2: 0.2913788828280721 runtime: 1.4998126029968262\n",
      "fold: 1\n",
      "train_r2: 0.9531617256890484 test_r2: -5.4825999932237846 independent_r2: 0.27842754678628223 runtime: 1.3081541061401367\n",
      "fold: 2\n",
      "train_r2: 0.9533250171169072 test_r2: -5.603517167631292 independent_r2: 0.39330460671501966 runtime: 1.5093116760253906\n",
      "fold: 3\n",
      "train_r2: 0.9531804448038979 test_r2: -6.440645858041065 independent_r2: 0.5171561139818579 runtime: 1.8427984714508057\n",
      "fold: 4\n",
      "train_r2: 0.9522870521167619 test_r2: -4.4605108592166545 independent_r2: 0.440716715357545 runtime: 1.5992496013641357\n",
      "fold: 5\n",
      "train_r2: 0.9540850407310756 test_r2: -7.140509497533987 independent_r2: 0.40493437713794467 runtime: 2.6211719512939453\n",
      "fold: 6\n",
      "train_r2: 0.9521088862768332 test_r2: -4.3396179246291755 independent_r2: 0.49167584800539554 runtime: 2.85577130317688\n",
      "fold: 7\n",
      "train_r2: 0.9517460586935949 test_r2: -3.9402069603018353 independent_r2: 0.3601782527699138 runtime: 2.31729793548584\n",
      "fold: 8\n",
      "train_r2: 0.9521148867918283 test_r2: -4.202792103195392 independent_r2: 0.3859461781817649 runtime: 1.351637840270996\n",
      "fold: 9\n",
      "train_r2: 0.9499994746958368 test_r2: -3.146294814083329 independent_r2: 0.2809991358924039 runtime: 2.422515869140625\n",
      "0.1 1\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9540667013534639 test_r2: -6.9131067166781435 independent_r2: 0.2913788819099795 runtime: 2.298462390899658\n",
      "fold: 1\n",
      "train_r2: 0.9531617246880064 test_r2: -5.4825999932237846 independent_r2: 0.27842754589496843 runtime: 1.3638300895690918\n",
      "fold: 2\n",
      "train_r2: 0.9533250160702927 test_r2: -5.603517167631292 independent_r2: 0.39330460575837545 runtime: 2.0302774906158447\n",
      "fold: 3\n",
      "train_r2: 0.9531804437388826 test_r2: -6.440645858041065 independent_r2: 0.5171561129879154 runtime: 3.1827337741851807\n",
      "fold: 4\n",
      "train_r2: 0.9522870509940669 test_r2: -4.4605108592166545 independent_r2: 0.44071671434436044 runtime: 2.644432783126831\n",
      "fold: 5\n",
      "train_r2: 0.9540850397272025 test_r2: -7.140509497533987 independent_r2: 0.4049343762044958 runtime: 2.075559139251709\n",
      "fold: 6\n",
      "train_r2: 0.9521088852493766 test_r2: -4.3396179246291755 independent_r2: 0.49167584707254697 runtime: 1.9718272686004639\n",
      "fold: 7\n",
      "train_r2: 0.9517460575534021 test_r2: -3.9402069603018353 independent_r2: 0.3601782517663613 runtime: 1.1942143440246582\n",
      "fold: 8\n",
      "train_r2: 0.9521148857156476 test_r2: -4.202792103195392 independent_r2: 0.38594617722236224 runtime: 2.4260807037353516\n",
      "fold: 9\n",
      "train_r2: 0.9499994734968759 test_r2: -3.146294814083329 independent_r2: 0.2809991348898312 runtime: 2.166731357574463\n",
      "1 1e-06\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.4181205640264686 test_r2: 0.21111844247236256 independent_r2: 0.40352937752989315 runtime: 2.0849335193634033\n",
      "fold: 1\n",
      "train_r2: 0.38464562335908425 test_r2: 0.423738480884319 independent_r2: 0.39132647450332203 runtime: 1.461033582687378\n",
      "fold: 2\n",
      "train_r2: 0.38751167114999685 test_r2: 0.4945891997748092 independent_r2: 0.3966750266081014 runtime: 1.6022753715515137\n",
      "fold: 3\n",
      "train_r2: 0.408329503726504 test_r2: 0.26524779769104445 independent_r2: 0.40414664101302866 runtime: 2.703073024749756\n",
      "fold: 4\n",
      "train_r2: 0.39534755150495327 test_r2: 0.40227804046768123 independent_r2: 0.3975269354816009 runtime: 3.4160678386688232\n",
      "fold: 5\n",
      "train_r2: 0.4043306178061259 test_r2: 0.4092682591985828 independent_r2: 0.40487197655705065 runtime: 2.0308640003204346\n",
      "fold: 6\n",
      "train_r2: 0.39139586869692844 test_r2: 0.41026081026358285 independent_r2: 0.39577412207159957 runtime: 1.4493908882141113\n",
      "fold: 7\n",
      "train_r2: 0.4120264933431804 test_r2: 0.273068911117875 independent_r2: 0.3953155420262623 runtime: 1.3364391326904297\n",
      "fold: 8\n",
      "train_r2: 0.39282042009440854 test_r2: 0.42054794892645797 independent_r2: 0.39586654503874785 runtime: 1.5629358291625977\n",
      "fold: 9\n",
      "train_r2: 0.4267368592677052 test_r2: 0.20833273402715458 independent_r2: 0.39200654716660965 runtime: 1.804305076599121\n",
      "1 1e-05\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.6939270544475538 test_r2: 0.3525931335700785 independent_r2: 0.6666650717533518 runtime: 1.1574206352233887\n",
      "fold: 1\n",
      "train_r2: 0.6724883341793675 test_r2: 0.6370876142386561 independent_r2: 0.670149543018674 runtime: 1.3460450172424316\n",
      "fold: 2\n",
      "train_r2: 0.6740637565629395 test_r2: 0.7158159411058345 independent_r2: 0.6776393124871447 runtime: 1.3826699256896973\n",
      "fold: 3\n",
      "train_r2: 0.6848180343061987 test_r2: 0.4759331209602965 independent_r2: 0.6747603078046074 runtime: 1.4987070560455322\n",
      "fold: 4\n",
      "train_r2: 0.6789510862780322 test_r2: 0.586512037650808 independent_r2: 0.6710218084320669 runtime: 1.400247573852539\n",
      "fold: 5\n",
      "train_r2: 0.6859583325591451 test_r2: 0.565376177684956 independent_r2: 0.6778864036383072 runtime: 2.1816365718841553\n",
      "fold: 6\n",
      "train_r2: 0.6792124577998313 test_r2: 0.5982281099728458 independent_r2: 0.6736048525923515 runtime: 1.4890973567962646\n",
      "fold: 7\n",
      "train_r2: 0.6885416118373731 test_r2: 0.45143188641571774 independent_r2: 0.6599175786440943 runtime: 1.7023584842681885\n",
      "fold: 8\n",
      "train_r2: 0.6849827705240785 test_r2: 0.57478027028806 independent_r2: 0.6728795214485588 runtime: 2.3462893962860107\n",
      "fold: 9\n",
      "train_r2: 0.7061277066646177 test_r2: 0.3005150098104118 independent_r2: 0.6403591944226591 runtime: 1.5462782382965088\n",
      "1 0.0001\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.6016467215457381 test_r2: -1.3179951262579372 independent_r2: 0.4418437358663898 runtime: 2.1390180587768555\n",
      "fold: 1\n",
      "train_r2: 0.5868030006197021 test_r2: -0.649726710687436 independent_r2: 0.45885892225649005 runtime: 3.175136089324951\n",
      "fold: 2\n",
      "train_r2: 0.5872910484418061 test_r2: -0.4497634142670701 independent_r2: 0.4987279460249793 runtime: 2.090723991394043\n",
      "fold: 3\n",
      "train_r2: 0.5984045534116562 test_r2: -0.6330992695549549 independent_r2: 0.5286172902575244 runtime: 2.9838051795959473\n",
      "fold: 4\n",
      "train_r2: 0.5967459574885416 test_r2: -0.6175182653479099 independent_r2: 0.48297386859051117 runtime: 3.0811767578125\n",
      "fold: 5\n",
      "train_r2: 0.5919999037725775 test_r2: -0.49611297835545076 independent_r2: 0.5183197429926074 runtime: 2.7286036014556885\n",
      "fold: 6\n",
      "train_r2: 0.5977027705759719 test_r2: -0.6083457427134673 independent_r2: 0.49452401742886654 runtime: 2.3055546283721924\n",
      "fold: 7\n",
      "train_r2: 0.5948596876690501 test_r2: -0.7420528109070101 independent_r2: 0.43325321321478705 runtime: 2.3274881839752197\n",
      "fold: 8\n",
      "train_r2: 0.5764050663283125 test_r2: -0.01762084967309918 independent_r2: 0.5111631305165936 runtime: 2.0346667766571045\n",
      "fold: 9\n",
      "train_r2: 0.5914054803153999 test_r2: -0.6037330633924811 independent_r2: 0.39686580100161495 runtime: 2.340334415435791\n",
      "1 0.001\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: -0.3894504402883743 test_r2: -6.912942287763633 independent_r2: -0.9323315720733754 runtime: 2.243621349334717\n",
      "fold: 1\n",
      "train_r2: -0.41684691474443314 test_r2: -5.48206063224445 independent_r2: -0.9421025918507608 runtime: 2.42120361328125\n",
      "fold: 2\n",
      "train_r2: -0.4119071771955918 test_r2: -5.602864909204658 independent_r2: -0.8552272554228224 runtime: 2.2860724925994873\n",
      "fold: 3\n",
      "train_r2: -0.41625113826477933 test_r2: -6.440643847399846 independent_r2: -0.7615925238435004 runtime: 1.8117778301239014\n",
      "fold: 4\n",
      "train_r2: -0.44327590018900564 test_r2: -4.460490185014333 independent_r2: -0.8193988980489344 runtime: 3.016366720199585\n",
      "fold: 5\n",
      "train_r2: -0.3888871821343898 test_r2: -7.140503389268851 independent_r2: -0.8464600618724276 runtime: 2.037376880645752\n",
      "fold: 6\n",
      "train_r2: -0.4486651713204093 test_r2: -4.339605785385367 independent_r2: -0.7808615037661724 runtime: 1.8603692054748535\n",
      "fold: 7\n",
      "train_r2: -0.4596387997438087 test_r2: -3.9402060190654744 independent_r2: -0.8803070398114552 runtime: 2.146618366241455\n",
      "fold: 8\n",
      "train_r2: -0.44849008470723994 test_r2: -4.202665329961835 independent_r2: -0.8608131726734862 runtime: 2.519765853881836\n",
      "fold: 9\n",
      "train_r2: -0.5124708451304476 test_r2: -3.146292627016794 independent_r2: -0.9401861171592947 runtime: 1.8033638000488281\n",
      "1 0.01\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: -0.3894822501925037 test_r2: -6.9131067166781435 independent_r2: -0.9323744009921067 runtime: 1.8712093830108643\n",
      "fold: 1\n",
      "train_r2: -0.4168577945753129 test_r2: -5.4825999932237846 independent_r2: -0.9421688488750604 runtime: 1.897841215133667\n",
      "fold: 2\n",
      "train_r2: -0.41191822873094375 test_r2: -5.603517167631292 independent_r2: -0.8552930723006591 runtime: 2.7490122318267822\n",
      "fold: 3\n",
      "train_r2: -0.41629154113824884 test_r2: -6.440645858041065 independent_r2: -0.7616303699282323 runtime: 2.7524518966674805\n",
      "fold: 4\n",
      "train_r2: -0.4433166697321813 test_r2: -4.4605108592166545 independent_r2: -0.8194376651368986 runtime: 2.1343677043914795\n",
      "fold: 5\n",
      "train_r2: -0.3889275145445734 test_r2: -7.140509497533987 independent_r2: -0.8464980584194912 runtime: 2.383082389831543\n",
      "fold: 6\n",
      "train_r2: -0.4487061867069313 test_r2: -4.3396179246291755 independent_r2: -0.7808998210760119 runtime: 2.3486127853393555\n",
      "fold: 7\n",
      "train_r2: -0.4596817207247619 test_r2: -3.9402069603018353 independent_r2: -0.8803448774692806 runtime: 1.7655012607574463\n",
      "fold: 8\n",
      "train_r2: -0.4485246709661974 test_r2: -4.202792103195392 independent_r2: -0.8608578838787937 runtime: 1.9967763423919678\n",
      "fold: 9\n",
      "train_r2: -0.512515886461391 test_r2: -3.146294814083329 independent_r2: -0.9402240846079466 runtime: 2.327029228210449\n",
      "1 0.1\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: -0.3894822504975961 test_r2: -6.9131067166781435 independent_r2: -0.9323744012698296 runtime: 2.3844363689422607\n",
      "fold: 1\n",
      "train_r2: -0.41685779487812846 test_r2: -5.4825999932237846 independent_r2: -0.9421688491446827 runtime: 2.5396790504455566\n",
      "fold: 2\n",
      "train_r2: -0.4119182290475447 test_r2: -5.603517167631292 independent_r2: -0.8552930725900438 runtime: 2.8046040534973145\n",
      "fold: 3\n",
      "train_r2: -0.416291541460416 test_r2: -6.440645858041065 independent_r2: -0.7616303702289 runtime: 2.298302173614502\n",
      "fold: 4\n",
      "train_r2: -0.44331667007179676 test_r2: -4.4605108592166545 independent_r2: -0.819437665443387 runtime: 2.0066003799438477\n",
      "fold: 5\n",
      "train_r2: -0.3889275148482454 test_r2: -7.140509497533987 independent_r2: -0.8464980587018598 runtime: 2.6606929302215576\n",
      "fold: 6\n",
      "train_r2: -0.44870618701773735 test_r2: -4.3396179246291755 independent_r2: -0.7808998213581988 runtime: 2.560243844985962\n",
      "fold: 7\n",
      "train_r2: -0.4596817210696702 test_r2: -3.9402069603018353 independent_r2: -0.8803448777728553 runtime: 1.931276798248291\n",
      "fold: 8\n",
      "train_r2: -0.4485246712917421 test_r2: -4.202792103195392 independent_r2: -0.8608578841690129 runtime: 1.6643495559692383\n",
      "fold: 9\n",
      "train_r2: -0.5125158868240767 test_r2: -3.146294814083329 independent_r2: -0.9402240849112251 runtime: 2.7032628059387207\n",
      "1 1\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: -0.3894822535485154 test_r2: -6.9131067166781435 independent_r2: -0.9323744040470605 runtime: 2.597637891769409\n",
      "fold: 1\n",
      "train_r2: -0.4168577979062802 test_r2: -5.4825999932237846 independent_r2: -0.942168851840907 runtime: 2.1164743900299072\n",
      "fold: 2\n",
      "train_r2: -0.4119182322135533 test_r2: -5.603517167631292 independent_r2: -0.8552930754838928 runtime: 2.7810463905334473\n",
      "fold: 3\n",
      "train_r2: -0.41629154468208696 test_r2: -6.440645858041065 independent_r2: -0.761630373235576 runtime: 2.8600549697875977\n",
      "fold: 4\n",
      "train_r2: -0.443316673467949 test_r2: -4.4605108592166545 independent_r2: -0.8194376685082703 runtime: 2.8932363986968994\n",
      "fold: 5\n",
      "train_r2: -0.3889275178849607 test_r2: -7.140509497533987 independent_r2: -0.8464980615255435 runtime: 1.7637183666229248\n",
      "fold: 6\n",
      "train_r2: -0.4487061901257927 test_r2: -4.3396179246291755 independent_r2: -0.7808998241800655 runtime: 1.5732216835021973\n",
      "fold: 7\n",
      "train_r2: -0.459681724518753 test_r2: -3.9402069603018353 independent_r2: -0.8803448808086014 runtime: 2.9524900913238525\n",
      "fold: 8\n",
      "train_r2: -0.4485246745471887 test_r2: -4.202792103195392 independent_r2: -0.8608578870712067 runtime: 2.1319832801818848\n",
      "fold: 9\n",
      "train_r2: -0.512515890450933 test_r2: -3.146294814083329 independent_r2: -0.9402240879440074 runtime: 2.256944417953491\n",
      "10 1e-06\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.0745948461895124 test_r2: -0.0823326368586299 independent_r2: 0.06590653525304424 runtime: 2.384363889694214\n",
      "fold: 1\n",
      "train_r2: 0.06482105956761075 test_r2: -0.011346315786309402 independent_r2: 0.06075585606141487 runtime: 2.201193332672119\n",
      "fold: 2\n",
      "train_r2: 0.06612365706084011 test_r2: 0.09400775514382764 independent_r2: 0.0685324363855524 runtime: 2.6366119384765625\n",
      "fold: 3\n",
      "train_r2: 0.07383259636603334 test_r2: 0.04694366149293705 independent_r2: 0.0789162330579759 runtime: 1.6745624542236328\n",
      "fold: 4\n",
      "train_r2: 0.06741233580639838 test_r2: 0.08985160832500028 independent_r2: 0.07188448509842382 runtime: 2.8056724071502686\n",
      "fold: 5\n",
      "train_r2: 0.07056608645648643 test_r2: 0.05354281049804388 independent_r2: 0.06973317078562602 runtime: 2.8963587284088135\n",
      "fold: 6\n",
      "train_r2: 0.06859209160781088 test_r2: 0.0794141312022385 independent_r2: 0.07372140535428895 runtime: 2.4865787029266357\n",
      "fold: 7\n",
      "train_r2: 0.07383979823033993 test_r2: 0.033829728159960104 independent_r2: 0.0691478606775695 runtime: 2.140725612640381\n",
      "fold: 8\n",
      "train_r2: 0.06533274891948493 test_r2: 0.08060577576952144 independent_r2: 0.06701140466585709 runtime: 1.916567325592041\n",
      "fold: 9\n",
      "train_r2: 0.07566016019398691 test_r2: 0.003589127560195138 independent_r2: 0.06540949086526882 runtime: 2.4523019790649414\n",
      "10 1e-05\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.31753606974725335 test_r2: 0.0846322404664559 independent_r2: 0.30125523200826654 runtime: 2.0914671421051025\n",
      "fold: 1\n",
      "train_r2: 0.2918135014193798 test_r2: 0.24576260179911924 independent_r2: 0.2899545488895543 runtime: 2.675424098968506\n",
      "fold: 2\n",
      "train_r2: 0.29467198220869417 test_r2: 0.38728679585799564 independent_r2: 0.30260277613560005 runtime: 3.036904811859131\n",
      "fold: 3\n",
      "train_r2: 0.31584157882309916 test_r2: 0.30799432445531816 independent_r2: 0.32030597580033093 runtime: 2.639289379119873\n",
      "fold: 4\n",
      "train_r2: 0.3023296585800287 test_r2: 0.3197148351168113 independent_r2: 0.3057318420175936 runtime: 1.9810152053833008\n",
      "fold: 5\n",
      "train_r2: 0.30691652718224127 test_r2: 0.3295665633543665 independent_r2: 0.30869330823957475 runtime: 1.529888391494751\n",
      "fold: 6\n",
      "train_r2: 0.30311237734357477 test_r2: 0.3205797227669476 independent_r2: 0.30776585731388284 runtime: 2.3975210189819336\n",
      "fold: 7\n",
      "train_r2: 0.3165716330958349 test_r2: 0.17979918972951403 independent_r2: 0.3001400167400332 runtime: 2.75827956199646\n",
      "fold: 8\n",
      "train_r2: 0.29592309362922353 test_r2: 0.34592381860530297 independent_r2: 0.3014156315166595 runtime: 2.288912057876587\n",
      "fold: 9\n",
      "train_r2: 0.3324835398173488 test_r2: 0.08823454356303362 independent_r2: 0.293686889770045 runtime: 1.8890447616577148\n",
      "10 0.0001\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: -1.892887969323084 test_r2: -4.097659369660377 independent_r2: -2.0644925481196976 runtime: 1.4410419464111328\n",
      "fold: 1\n",
      "train_r2: -1.9767478135731338 test_r2: -3.012861440945861 independent_r2: -2.0729210413597015 runtime: 3.2311482429504395\n",
      "fold: 2\n",
      "train_r2: -1.9473587110349535 test_r2: -3.0105155295262946 independent_r2: -2.038077424883225 runtime: 1.7948107719421387\n",
      "fold: 3\n",
      "train_r2: -1.9421997679827276 test_r2: -3.243521526519591 independent_r2: -1.997809266681628 runtime: 3.0422861576080322\n",
      "fold: 4\n",
      "train_r2: -1.9764722635998684 test_r2: -2.4477472874136597 independent_r2: -2.01352037614384 runtime: 2.1674420833587646\n",
      "fold: 5\n",
      "train_r2: -1.9236578205000443 test_r2: -3.630798285271264 independent_r2: -2.0384634949036537 runtime: 2.244884967803955\n",
      "fold: 6\n",
      "train_r2: -1.9801183579480393 test_r2: -2.337199532320042 independent_r2: -1.9978050576143098 runtime: 2.0229032039642334\n",
      "fold: 7\n",
      "train_r2: -1.9919365950551504 test_r2: -2.3660747887818245 independent_r2: -2.03670750353169 runtime: 3.078333854675293\n",
      "fold: 8\n",
      "train_r2: -2.0429257036804396 test_r2: -2.0500973568426213 independent_r2: -2.043709455954207 runtime: 1.714695930480957\n",
      "fold: 9\n",
      "train_r2: -2.09429090825618 test_r2: -1.8739061038504814 independent_r2: -2.053200904612707 runtime: 2.1366705894470215\n",
      "10 0.001\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: -3.59331047893967 test_r2: -6.913076819501587 independent_r2: -3.850535234328899 runtime: 1.1385765075683594\n",
      "fold: 1\n",
      "train_r2: -3.683820879197304 test_r2: -5.482501646014762 independent_r2: -3.852805651253794 runtime: 2.8512237071990967\n",
      "fold: 2\n",
      "train_r2: -3.6674916327122595 test_r2: -5.603398253100674 independent_r2: -3.8327020741857147 runtime: 2.188119649887085\n",
      "fold: 3\n",
      "train_r2: -3.681931072360662 test_r2: -6.4406454924697165 independent_r2: -3.8110214543543632 runtime: 2.243788480758667\n",
      "fold: 4\n",
      "train_r2: -3.7712701155779778 test_r2: -4.460507100201805 independent_r2: -3.824398565768793 runtime: 2.7526159286499023\n",
      "fold: 5\n",
      "train_r2: -3.591471522358553 test_r2: -7.1405083869162755 independent_r2: -3.8306602918557937 runtime: 2.0074284076690674\n",
      "fold: 6\n",
      "train_r2: -3.7890865531215363 test_r2: -4.339615717478146 independent_r2: -3.8154806109426707 runtime: 1.5714857578277588\n",
      "fold: 7\n",
      "train_r2: -3.8253681602924523 test_r2: -3.940206787150462 independent_r2: -3.8384925834633288 runtime: 2.603508472442627\n",
      "fold: 8\n",
      "train_r2: -3.7884903628484885 test_r2: -4.202769051898245 independent_r2: -3.83398485480454 runtime: 2.0373291969299316\n",
      "fold: 9\n",
      "train_r2: -4.000025275847559 test_r2: -3.1462944164247126 independent_r2: -3.8523489525936068 runtime: 2.2785086631774902\n",
      "10 0.01\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: -3.5933297526016332 test_r2: -6.9131067166781435 independent_r2: -3.850555308712315 runtime: 2.5255253314971924\n",
      "fold: 1\n",
      "train_r2: -3.683827419983573 test_r2: -5.4825999932237846 independent_r2: -3.852821792519852 runtime: 1.8120582103729248\n",
      "fold: 2\n",
      "train_r2: -3.6674982766918447 test_r2: -5.603517167631292 independent_r2: -3.8327183070317776 runtime: 2.7609481811523438\n",
      "fold: 3\n",
      "train_r2: -3.6819555077885333 test_r2: -6.440645858041065 independent_r2: -3.81104429325977 runtime: 3.3422951698303223\n",
      "fold: 4\n",
      "train_r2: -3.7712947758618816 test_r2: -4.4605108592166545 independent_r2: -3.8244211880188645 runtime: 1.8862154483795166\n",
      "fold: 5\n",
      "train_r2: -3.591495915749441 test_r2: -7.140509497533987 independent_r2: -3.8306830972079577 runtime: 2.3720955848693848\n",
      "fold: 6\n",
      "train_r2: -3.7891113609119014 test_r2: -4.3396179246291755 independent_r2: -3.8155033398061606 runtime: 2.7498838901519775\n",
      "fold: 7\n",
      "train_r2: -3.8253941179843647 test_r2: -3.9402069603018353 independent_r2: -3.83851541897178 runtime: 2.403533458709717\n",
      "fold: 8\n",
      "train_r2: -3.788511308871545 test_r2: -4.202792103195392 independent_r2: -3.8340060320250737 runtime: 2.363389730453491\n",
      "fold: 9\n",
      "train_r2: -4.00005251710784 test_r2: -3.146294814083329 independent_r2: -3.852371764425352 runtime: 2.2663683891296387\n",
      "10 0.1\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: -3.5933297527024903 test_r2: -6.9131067166781435 independent_r2: -3.8505553088041244 runtime: 3.167809247970581\n",
      "fold: 1\n",
      "train_r2: -3.683827420083677 test_r2: -5.4825999932237846 independent_r2: -3.8528217926089825 runtime: 1.5053489208221436\n",
      "fold: 2\n",
      "train_r2: -3.6674982767965068 test_r2: -5.603517167631292 independent_r2: -3.8327183071274424 runtime: 1.3747310638427734\n",
      "fold: 3\n",
      "train_r2: -3.6819555078950366 test_r2: -6.440645858041065 independent_r2: -3.8110442933591653 runtime: 1.466456413269043\n",
      "fold: 4\n",
      "train_r2: -3.7712947759741517 test_r2: -4.4605108592166545 independent_r2: -3.8244211881201835 runtime: 1.4272606372833252\n",
      "fold: 5\n",
      "train_r2: -3.591495915849827 test_r2: -7.140509497533987 independent_r2: -3.8306830973013044 runtime: 1.9641263484954834\n",
      "fold: 6\n",
      "train_r2: -3.7891113610146476 test_r2: -4.3396179246291755 independent_r2: -3.815503339899445 runtime: 2.1300270557403564\n",
      "fold: 7\n",
      "train_r2: -3.8253941180983837 test_r2: -3.9402069603018353 independent_r2: -3.8385154190721362 runtime: 2.6792070865631104\n",
      "fold: 8\n",
      "train_r2: -3.7885113089791638 test_r2: -4.202792103195392 independent_r2: -3.8340060321210157 runtime: 2.034281015396118\n",
      "fold: 9\n",
      "train_r2: -4.000052517227737 test_r2: -3.146294814083329 independent_r2: -3.8523717645256106 runtime: 2.1445937156677246\n",
      "10 1\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: -3.5933297537110587 test_r2: -6.9131067166781435 independent_r2: -3.8505553097222176 runtime: 1.8996765613555908\n",
      "fold: 1\n",
      "train_r2: -3.68382742108472 test_r2: -5.4825999932237846 independent_r2: -3.852821793500296 runtime: 2.691204786300659\n",
      "fold: 2\n",
      "train_r2: -3.667498277843121 test_r2: -5.603517167631292 independent_r2: -3.832718308084086 runtime: 1.8778297901153564\n",
      "fold: 3\n",
      "train_r2: -3.6819555089600504 test_r2: -6.440645858041065 independent_r2: -3.811044294353107 runtime: 2.9216978549957275\n",
      "fold: 4\n",
      "train_r2: -3.7712947770968466 test_r2: -4.4605108592166545 independent_r2: -3.824421189133367 runtime: 2.503528594970703\n",
      "fold: 5\n",
      "train_r2: -3.5914959168537015 test_r2: -7.140509497533987 independent_r2: -3.8306830982347533 runtime: 2.614359140396118\n",
      "fold: 6\n",
      "train_r2: -3.789111362042104 test_r2: -4.3396179246291755 independent_r2: -3.8155033408322936 runtime: 1.9567618370056152\n",
      "fold: 7\n",
      "train_r2: -3.8253941192385765 test_r2: -3.9402069603018353 independent_r2: -3.838515420075689 runtime: 2.4768564701080322\n",
      "fold: 8\n",
      "train_r2: -3.788511310055344 test_r2: -4.202792103195392 independent_r2: -3.834006033080418 runtime: 2.5719408988952637\n",
      "fold: 9\n",
      "train_r2: -4.000052518426697 test_r2: -3.146294814083329 independent_r2: -3.852371765528183 runtime: 2.9530413150787354\n"
     ]
    }
   ],
   "source": [
    "import itertools\n",
    "feature_sets = ['Expr']\n",
    "methods = ['KernelRidge']\n",
    "targets = ['auc']\n",
    "# TODO: rbf kernel bandwith? median\n",
    "alpha_vals = [1e-3, 1e-2, 1e-1, 1, 10]\n",
    "gam_vals = [1e-6, 1e-5, 1e-4, 1e-3, 1e-2, 1e-1, 1]\n",
    "results_vals = {}\n",
    "for dataset, method, target, alpha_val, gam_val in itertools.product(feature_sets, methods, targets, alpha_vals, gam_vals):\n",
    "    print(alpha_val, gam_val)\n",
    "    result_all = run_regression_dataset(dataset, dic_dataset_train, dic_dataset_train, method=method, target=target,\n",
    "                                       model_kwargs={'kernel': 'rbf', 'alpha':alpha_val, 'gamma': gam_val})\n",
    "    results_vals[alpha_val, gam_val] = result_all"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 272,
   "metadata": {},
   "outputs": [],
   "source": [
    "# figure out the best hyperparameters for both expr and relativeexpr\n",
    "mean_test_r2 = {k: np.mean([x['test_r2'] for iter, x in row.items()]) for k, row in results_vals.items()}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 281,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[((0.1, 1e-05), 0.5517300498323034),\n",
       " ((0.01, 1e-06), 0.5456347752765717),\n",
       " ((0.1, 1e-06), 0.5378786096943446),\n",
       " ((1, 1e-05), 0.5258273301697665),\n",
       " ((0.01, 1e-05), 0.5237893209008141),\n",
       " ((0.001, 1e-05), 0.5160553172933262),\n",
       " ((0.001, 1e-06), 0.4990796909102113),\n",
       " ((1, 1e-06), 0.35184506248238695),\n",
       " ((10, 1e-05), 0.2609494635714865),\n",
       " ((10, 1e-06), 0.038810564550678474),\n",
       " ((0.001, 0.0001), -0.1278957195630009),\n",
       " ((0.01, 0.0001), -0.13258142242417942),\n",
       " ((0.1, 0.0001), -0.1792121015123984),\n",
       " ((1, 0.0001), -0.6135968231156816),\n",
       " ((10, 0.0001), -2.8070381221132013),\n",
       " ((0.001, 0.001), -5.166675933634637),\n",
       " ((0.01, 0.001), -5.166678630405615),\n",
       " ((0.1, 0.001), -5.16670318293726),\n",
       " ((1, 0.001), -5.1668275012325235),\n",
       " ((10, 0.001), -5.1669523671156385),\n",
       " ((10, 1), -5.166980189453465),\n",
       " ((10, 0.1), -5.166980189453465),\n",
       " ((10, 0.01), -5.166980189453465),\n",
       " ((1, 1), -5.166980189453465),\n",
       " ((1, 0.1), -5.166980189453465),\n",
       " ((1, 0.01), -5.166980189453465),\n",
       " ((0.1, 1), -5.166980189453465),\n",
       " ((0.1, 0.1), -5.166980189453465),\n",
       " ((0.1, 0.01), -5.166980189453465),\n",
       " ((0.01, 1), -5.166980189453465),\n",
       " ((0.01, 0.1), -5.166980189453465),\n",
       " ((0.01, 0.01), -5.166980189453465),\n",
       " ((0.001, 1), -5.166980189453465),\n",
       " ((0.001, 0.1), -5.166980189453465),\n",
       " ((0.001, 0.01), -5.166980189453465)]"
      ]
     },
     "execution_count": 281,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted(list(mean_test_r2.items()), key=lambda x: x[1])[::-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 274,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.001 1e-06\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.6502416536078504 test_r2: 0.36988879720936463 independent_r2: 0.6283322060763985 runtime: 0.06204342842102051\n",
      "fold: 1\n",
      "train_r2: 0.6081567588630349 test_r2: 0.725360243625043 independent_r2: 0.6220917018241698 runtime: 0.06093096733093262\n",
      "fold: 2\n",
      "train_r2: 0.6224907888749303 test_r2: 0.6635912493018163 independent_r2: 0.6260121824419556 runtime: 0.08225131034851074\n",
      "fold: 3\n",
      "train_r2: 0.6283146028009617 test_r2: 0.5588015505599313 independent_r2: 0.6268891421922506 runtime: 0.04103565216064453\n",
      "fold: 4\n",
      "train_r2: 0.6209384071414801 test_r2: 0.6540312624824144 independent_r2: 0.6250223363798262 runtime: 0.042041778564453125\n",
      "fold: 5\n",
      "train_r2: 0.6331287586648247 test_r2: 0.6159642700231764 independent_r2: 0.6320913612259557 runtime: 0.04124021530151367\n",
      "fold: 6\n",
      "train_r2: 0.6168109592557179 test_r2: 0.6754589165480924 independent_r2: 0.6236389719578417 runtime: 0.04284405708312988\n",
      "fold: 7\n",
      "train_r2: 0.6341434616311314 test_r2: 0.5066918335826314 independent_r2: 0.6187887923413281 runtime: 0.04357266426086426\n",
      "fold: 8\n",
      "train_r2: 0.6224807806899051 test_r2: 0.6505501148544544 independent_r2: 0.625564150776152 runtime: 0.03997993469238281\n",
      "fold: 9\n",
      "train_r2: 0.6534092361100019 test_r2: 0.4429872208957777 independent_r2: 0.6196096458272786 runtime: 0.041565656661987305\n",
      "0.001 1e-05\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.8241893784669176 test_r2: 0.38963952785382583 independent_r2: 0.7884343836517586 runtime: 0.04526329040527344\n",
      "fold: 1\n",
      "train_r2: 0.8035415409613847 test_r2: 0.7945430221256349 independent_r2: 0.8034218955088286 runtime: 0.03872537612915039\n",
      "fold: 2\n",
      "train_r2: 0.8087585861318284 test_r2: 0.7452594706354047 independent_r2: 0.8033406766253739 runtime: 0.04130816459655762\n",
      "fold: 3\n",
      "train_r2: 0.8056196885366911 test_r2: 0.7269818600667919 independent_r2: 0.8023787053642654 runtime: 0.03649568557739258\n",
      "fold: 4\n",
      "train_r2: 0.810872185881448 test_r2: 0.7158789300867338 independent_r2: 0.8023689004351514 runtime: 0.033408403396606445\n",
      "fold: 5\n",
      "train_r2: 0.8170325928748376 test_r2: 0.5878530980427532 independent_r2: 0.8015476165847921 runtime: 0.043511152267456055\n",
      "fold: 6\n",
      "train_r2: 0.8064825196136692 test_r2: 0.7752805180205711 independent_r2: 0.8046363440135367 runtime: 0.043248891830444336\n",
      "fold: 7\n",
      "train_r2: 0.8138886561773329 test_r2: 0.6309472192942014 independent_r2: 0.7917952852433894 runtime: 0.062444448471069336\n",
      "fold: 8\n",
      "train_r2: 0.8105745881948897 test_r2: 0.6814948491454991 independent_r2: 0.7963978707401724 runtime: 0.03493618965148926\n",
      "fold: 9\n",
      "train_r2: 0.8279924628170255 test_r2: 0.5137669204722896 independent_r2: 0.77695042412788 runtime: 0.04293203353881836\n",
      "0.001 0.0001\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9556320120622922 test_r2: 0.28625515274143987 independent_r2: 0.8994456237126325 runtime: 0.04219508171081543\n",
      "fold: 1\n",
      "train_r2: 0.9527779464008014 test_r2: 0.608440225318233 independent_r2: 0.9168646641559915 runtime: 0.04115772247314453\n",
      "fold: 2\n",
      "train_r2: 0.9525906220436702 test_r2: 0.6778730092149087 independent_r2: 0.9291282923919383 runtime: 0.04689788818359375\n",
      "fold: 3\n",
      "train_r2: 0.9480345884350436 test_r2: 0.8048616050137715 independent_r2: 0.93995921748734 runtime: 0.04487276077270508\n",
      "fold: 4\n",
      "train_r2: 0.9552846075165446 test_r2: 0.5971755698815472 independent_r2: 0.9215440341778665 runtime: 0.045217275619506836\n",
      "fold: 5\n",
      "train_r2: 0.9525265387352223 test_r2: 0.47677691529688593 independent_r2: 0.9202664108455119 runtime: 0.03344249725341797\n",
      "fold: 6\n",
      "train_r2: 0.9529276723603352 test_r2: 0.6856499683237187 independent_r2: 0.9298725893655091 runtime: 0.04112052917480469\n",
      "fold: 7\n",
      "train_r2: 0.9515334936124106 test_r2: 0.6722701891404683 independent_r2: 0.9177703260483433 runtime: 0.04792523384094238\n",
      "fold: 8\n",
      "train_r2: 0.952103216647044 test_r2: 0.640084532834285 independent_r2: 0.9178339433189129 runtime: 0.035341501235961914\n",
      "fold: 9\n",
      "train_r2: 0.9547586606820028 test_r2: 0.5703348833125272 independent_r2: 0.8920419108351573 runtime: 0.037362098693847656\n",
      "0.001 0.001\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9980767004539269 test_r2: 0.15340201510143903 independent_r2: 0.9269111651149525 runtime: 0.05269503593444824\n",
      "fold: 1\n",
      "train_r2: 0.9980056019420356 test_r2: 0.4382358478539262 independent_r2: 0.9393098425684505 runtime: 0.04100799560546875\n",
      "fold: 2\n",
      "train_r2: 0.9979765024121282 test_r2: 0.5751066291629495 independent_r2: 0.9618591243113674 runtime: 0.04417610168457031\n",
      "fold: 3\n",
      "train_r2: 0.9978512064469753 test_r2: 0.6827579523369595 independent_r2: 0.9792708201886884 runtime: 0.05817556381225586\n",
      "fold: 4\n",
      "train_r2: 0.9982036958584583 test_r2: 0.44268330305431625 independent_r2: 0.945692949157292 runtime: 0.04831242561340332\n",
      "fold: 5\n",
      "train_r2: 0.9978072283066105 test_r2: 0.39347469288682546 independent_r2: 0.9568078829745348 runtime: 0.056733131408691406\n",
      "fold: 6\n",
      "train_r2: 0.9980865812784672 test_r2: 0.5637506467368751 independent_r2: 0.9602859513121526 runtime: 0.06222343444824219\n",
      "fold: 7\n",
      "train_r2: 0.997970847514479 test_r2: 0.6569852291529081 independent_r2: 0.9567363660717408 runtime: 0.05045580863952637\n",
      "fold: 8\n",
      "train_r2: 0.9980568125495756 test_r2: 0.5903444433548032 independent_r2: 0.9532773435428012 runtime: 0.043013811111450195\n",
      "fold: 9\n",
      "train_r2: 0.998077700530497 test_r2: 0.5321635749896688 independent_r2: 0.9219791423526479 runtime: 0.038860321044921875\n",
      "0.001 0.01\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9999943728402425 test_r2: 0.41410498682847197 independent_r2: 0.9506255078149225 runtime: 0.04004311561584473\n",
      "fold: 1\n",
      "train_r2: 0.9999943226009551 test_r2: 0.6837195862141305 independent_r2: 0.9668260098103221 runtime: 0.055039167404174805\n",
      "fold: 2\n",
      "train_r2: 0.9999942792526537 test_r2: 0.7203045035610856 independent_r2: 0.9761058955528956 runtime: 0.037079572677612305\n",
      "fold: 3\n",
      "train_r2: 0.99999385033931 test_r2: 0.762628556488617 independent_r2: 0.9859853041084431 runtime: 0.03607463836669922\n",
      "fold: 4\n",
      "train_r2: 0.9999944480130872 test_r2: 0.6395325738160221 independent_r2: 0.9659187473341985 runtime: 0.047101736068725586\n",
      "fold: 5\n",
      "train_r2: 0.9999946367009958 test_r2: 0.4322363701418749 independent_r2: 0.9614758642208718 runtime: 0.04005146026611328\n",
      "fold: 6\n",
      "train_r2: 0.9999941642211087 test_r2: 0.7368129691951206 independent_r2: 0.9770840924030451 runtime: 0.04627370834350586\n",
      "fold: 7\n",
      "train_r2: 0.9999940827790313 test_r2: 0.6865989753459418 independent_r2: 0.9620957543549336 runtime: 0.036405086517333984\n",
      "fold: 8\n",
      "train_r2: 0.9999941265185742 test_r2: 0.6641516729817847 independent_r2: 0.9631082022426752 runtime: 0.033399343490600586\n",
      "fold: 9\n",
      "train_r2: 0.9999946167028932 test_r2: 0.5075881534112965 independent_r2: 0.919565684467327 runtime: 0.03890872001647949\n",
      "0.001 0.1\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9999964592589965 test_r2: -4.069461499672766 independent_r2: 0.5728279163075409 runtime: 0.054535865783691406\n",
      "fold: 1\n",
      "train_r2: 0.9999963902022437 test_r2: -3.1623511818575416 independent_r2: 0.5634830533373216 runtime: 0.047893524169921875\n",
      "fold: 2\n",
      "train_r2: 0.9999963888937798 test_r2: -2.9330555172418933 independent_r2: 0.6640732989695299 runtime: 0.044945716857910156\n",
      "fold: 3\n",
      "train_r2: 0.9999964233472416 test_r2: -3.5100877969151245 independent_r2: 0.733824752145381 runtime: 0.05858969688415527\n",
      "fold: 4\n",
      "train_r2: 0.9999963588505536 test_r2: -2.4501455518112043 independent_r2: 0.6738424534273952 runtime: 0.03399467468261719\n",
      "fold: 5\n",
      "train_r2: 0.999996438818927 test_r2: -3.714917718841937 independent_r2: 0.6801196101408611 runtime: 0.03896450996398926\n",
      "fold: 6\n",
      "train_r2: 0.9999963740759827 test_r2: -2.454268723057403 independent_r2: 0.6993003164412477 runtime: 0.04551506042480469\n",
      "fold: 7\n",
      "train_r2: 0.9999962767612851 test_r2: -2.078716644128729 independent_r2: 0.6276928382982604 runtime: 0.03860759735107422\n",
      "fold: 8\n",
      "train_r2: 0.9999962772569232 test_r2: -1.9530598827582915 independent_r2: 0.6756590626659523 runtime: 0.04566192626953125\n",
      "fold: 9\n",
      "train_r2: 0.9999962660168245 test_r2: -2.0139515505734717 independent_r2: 0.5077025078092054 runtime: 0.05329418182373047\n",
      "0.001 1\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9999944531702112 test_r2: -6.913106715593374 independent_r2: 0.3332115498744389 runtime: 0.07002067565917969\n",
      "fold: 1\n",
      "train_r2: 0.9999943438867048 test_r2: -5.4825999697508925 independent_r2: 0.320152301432721 runtime: 0.09573769569396973\n",
      "fold: 2\n",
      "train_r2: 0.9999943636055109 test_r2: -5.603517138824771 independent_r2: 0.4359865545971964 runtime: 0.07204484939575195\n",
      "fold: 3\n",
      "train_r2: 0.9999943461472005 test_r2: -6.440645858040477 independent_r2: 0.56087003066756 runtime: 0.06250333786010742\n",
      "fold: 4\n",
      "train_r2: 0.9999942382625635 test_r2: -4.460510859214677 independent_r2: 0.4837937139794909 runtime: 0.06446218490600586\n",
      "fold: 5\n",
      "train_r2: 0.9999944553847226 test_r2: -7.140509497347175 independent_r2: 0.447713225620086 runtime: 0.07027435302734375\n",
      "fold: 6\n",
      "train_r2: 0.9999942167475471 test_r2: -4.339617924627394 independent_r2: 0.5351774547758628 runtime: 0.05962371826171875\n",
      "fold: 7\n",
      "train_r2: 0.999994172933084 test_r2: -3.940206960301807 independent_r2: 0.40258417858314055 runtime: 0.04584360122680664\n",
      "fold: 8\n",
      "train_r2: 0.99999421747216 test_r2: -4.202792102245041 independent_r2: 0.42856681087541504 runtime: 0.04267740249633789\n",
      "fold: 9\n",
      "train_r2: 0.9999939620184615 test_r2: -3.1462948140682725 independent_r2: 0.3227453154135017 runtime: 0.05719137191772461\n",
      "0.01 1e-06\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.2666286253832326 test_r2: 0.1467555415092724 independent_r2: 0.2601215847461993 runtime: 0.05447816848754883\n",
      "fold: 1\n",
      "train_r2: 0.24147628596736004 test_r2: 0.264715625546882 independent_r2: 0.24709507318245916 runtime: 0.04109025001525879\n",
      "fold: 2\n",
      "train_r2: 0.25372353070066966 test_r2: 0.2716276445866874 independent_r2: 0.25527445460319587 runtime: 0.03574490547180176\n",
      "fold: 3\n",
      "train_r2: 0.2613681663183458 test_r2: 0.1474487019275037 independent_r2: 0.25996483254744307 runtime: 0.03447222709655762\n",
      "fold: 4\n",
      "train_r2: 0.25014627282069124 test_r2: 0.25328338094952374 independent_r2: 0.25233308178457825 runtime: 0.038160085678100586\n",
      "fold: 5\n",
      "train_r2: 0.25572115151085406 test_r2: 0.29544247136100354 independent_r2: 0.25867384697739904 runtime: 0.037824153900146484\n",
      "fold: 6\n",
      "train_r2: 0.25170854744152216 test_r2: 0.23103615895297624 independent_r2: 0.25327302985356626 runtime: 0.0448460578918457\n",
      "fold: 7\n",
      "train_r2: 0.2636166715217716 test_r2: 0.18729970667142426 independent_r2: 0.2545042031594924 runtime: 0.03853034973144531\n",
      "fold: 8\n",
      "train_r2: 0.2450468877374139 test_r2: 0.2838711910838815 independent_r2: 0.2493119731731367 runtime: 0.04099917411804199\n",
      "fold: 9\n",
      "train_r2: 0.26013105973580064 test_r2: 0.1796804947818914 independent_r2: 0.2482080822973447 runtime: 0.041001319885253906\n",
      "0.01 1e-05\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.6502023034918727 test_r2: 0.36990028827389654 independent_r2: 0.6282973328321229 runtime: 0.03732705116271973\n",
      "fold: 1\n",
      "train_r2: 0.6081242997434392 test_r2: 0.7252277941855784 independent_r2: 0.6220488926784422 runtime: 0.09119176864624023\n",
      "fold: 2\n",
      "train_r2: 0.6224562146988792 test_r2: 0.6635826312274479 independent_r2: 0.625979826190499 runtime: 0.05810046195983887\n",
      "fold: 3\n",
      "train_r2: 0.6282851861895888 test_r2: 0.5587998505466216 independent_r2: 0.6268615731991578 runtime: 0.059578657150268555\n",
      "fold: 4\n",
      "train_r2: 0.6209028970582737 test_r2: 0.6540292052217198 independent_r2: 0.624990078222871 runtime: 0.052248239517211914\n",
      "fold: 5\n",
      "train_r2: 0.6330932801357334 test_r2: 0.6159740287369919 independent_r2: 0.632058964046638 runtime: 0.05914616584777832\n",
      "fold: 6\n",
      "train_r2: 0.6167770017966059 test_r2: 0.6754433712137761 independent_r2: 0.6236067699459729 runtime: 0.043615102767944336\n",
      "fold: 7\n",
      "train_r2: 0.6341074145804186 test_r2: 0.5066719283437715 independent_r2: 0.6187547029905589 runtime: 0.03891181945800781\n",
      "fold: 8\n",
      "train_r2: 0.6224446074216337 test_r2: 0.6505238661191897 independent_r2: 0.6255290675762999 runtime: 0.04033637046813965\n",
      "fold: 9\n",
      "train_r2: 0.6533754192555268 test_r2: 0.44291095278868164 independent_r2: 0.6195689507049879 runtime: 0.03853487968444824\n",
      "0.01 0.0001\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.824284414537786 test_r2: 0.3902355887299713 independent_r2: 0.788571171938379 runtime: 0.04546952247619629\n",
      "fold: 1\n",
      "train_r2: 0.8036667024951114 test_r2: 0.7945859306137516 independent_r2: 0.8035379060174834 runtime: 0.04076886177062988\n",
      "fold: 2\n",
      "train_r2: 0.808908183417763 test_r2: 0.7452551485293493 independent_r2: 0.8034771232562843 runtime: 0.044184207916259766\n",
      "fold: 3\n",
      "train_r2: 0.8058101580401456 test_r2: 0.7262801044867624 independent_r2: 0.8025151465889938 runtime: 0.04114794731140137\n",
      "fold: 4\n",
      "train_r2: 0.8110090264580948 test_r2: 0.7160449624298346 independent_r2: 0.8025081556602456 runtime: 0.04418134689331055\n",
      "fold: 5\n",
      "train_r2: 0.817197434199859 test_r2: 0.5876564073880615 independent_r2: 0.8016878731629102 runtime: 0.057256221771240234\n",
      "fold: 6\n",
      "train_r2: 0.8066171524946856 test_r2: 0.7754123859926679 independent_r2: 0.804770130950289 runtime: 0.04168200492858887\n",
      "fold: 7\n",
      "train_r2: 0.8140249382139195 test_r2: 0.6309407054275304 independent_r2: 0.7919142776738457 runtime: 0.052187204360961914\n",
      "fold: 8\n",
      "train_r2: 0.8106931817430514 test_r2: 0.6816048946793343 independent_r2: 0.7965155252984626 runtime: 0.06339526176452637\n",
      "fold: 9\n",
      "train_r2: 0.828159014115552 test_r2: 0.5131967250994203 independent_r2: 0.7769963622031252 runtime: 0.07972955703735352\n",
      "0.01 0.001\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.961957829611386 test_r2: 0.3197298402362977 independent_r2: 0.9080280915969631 runtime: 0.06424593925476074\n",
      "fold: 1\n",
      "train_r2: 0.9600697121668946 test_r2: 0.6377674506302894 independent_r2: 0.926436745351779 runtime: 0.06095385551452637\n",
      "fold: 2\n",
      "train_r2: 0.9598124887616054 test_r2: 0.6933769245931202 independent_r2: 0.9370573209884151 runtime: 0.07789063453674316\n",
      "fold: 3\n",
      "train_r2: 0.9561441262566395 test_r2: 0.8050345216192496 independent_r2: 0.9475419522600116 runtime: 0.04551291465759277\n",
      "fold: 4\n",
      "train_r2: 0.9615950235787796 test_r2: 0.6174238512629757 independent_r2: 0.9291561376943955 runtime: 0.05649542808532715\n",
      "fold: 5\n",
      "train_r2: 0.9603632012320219 test_r2: 0.47525116332219375 independent_r2: 0.9274651782481553 runtime: 0.049141883850097656\n",
      "fold: 6\n",
      "train_r2: 0.9598498215563472 test_r2: 0.7017640530588218 independent_r2: 0.937563783370428 runtime: 0.05702018737792969\n",
      "fold: 7\n",
      "train_r2: 0.9589266641407906 test_r2: 0.6734723795058335 independent_r2: 0.9244136632901001 runtime: 0.04269552230834961\n",
      "fold: 8\n",
      "train_r2: 0.9590290088166215 test_r2: 0.6518795850769721 independent_r2: 0.9252945224451254 runtime: 0.04847145080566406\n",
      "fold: 9\n",
      "train_r2: 0.9616700179652521 test_r2: 0.5660978789474185 independent_r2: 0.8971209363495275 runtime: 0.05375552177429199\n",
      "0.01 0.01\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9994924771648189 test_r2: 0.41692939862122813 independent_r2: 0.95040635683764 runtime: 0.054735660552978516\n",
      "fold: 1\n",
      "train_r2: 0.9994853580945615 test_r2: 0.6907073128932764 independent_r2: 0.9671053725042348 runtime: 0.05220484733581543\n",
      "fold: 2\n",
      "train_r2: 0.9994826633618483 test_r2: 0.7239726022932891 independent_r2: 0.9759512851815161 runtime: 0.047153472900390625\n",
      "fold: 3\n",
      "train_r2: 0.9994458141526625 test_r2: 0.761235122781324 independent_r2: 0.9853913222101859 runtime: 0.04839324951171875\n",
      "fold: 4\n",
      "train_r2: 0.9994978372005291 test_r2: 0.6448880545595694 independent_r2: 0.9659766069574384 runtime: 0.03864645957946777\n",
      "fold: 5\n",
      "train_r2: 0.9995145644452712 test_r2: 0.44093829294972875 independent_r2: 0.9616188980431368 runtime: 0.06663155555725098\n",
      "fold: 6\n",
      "train_r2: 0.9994725168331039 test_r2: 0.7410678546660814 independent_r2: 0.9769805905009459 runtime: 0.05126023292541504\n",
      "fold: 7\n",
      "train_r2: 0.9994664910803991 test_r2: 0.6843893144965039 independent_r2: 0.9613648363413129 runtime: 0.04797697067260742\n",
      "fold: 8\n",
      "train_r2: 0.9994689003657263 test_r2: 0.6656892563524819 independent_r2: 0.9628095372711329 runtime: 0.044937849044799805\n",
      "fold: 9\n",
      "train_r2: 0.9995146739825584 test_r2: 0.5065835453397283 independent_r2: 0.9190008335911148 runtime: 0.061011552810668945\n",
      "0.01 0.1\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9996511157560533 test_r2: -4.0859086726439005 independent_r2: 0.5711274740660188 runtime: 0.057161808013916016\n",
      "fold: 1\n",
      "train_r2: 0.9996443055012053 test_r2: -3.1754919267621124 independent_r2: 0.5617912741992355 runtime: 0.04268002510070801\n",
      "fold: 2\n",
      "train_r2: 0.9996441929439416 test_r2: -2.9481070801180755 independent_r2: 0.6624656361339443 runtime: 0.03857898712158203\n",
      "fold: 3\n",
      "train_r2: 0.9996475628180478 test_r2: -3.526671419761228 independent_r2: 0.7325202770454011 runtime: 0.04047274589538574\n",
      "fold: 4\n",
      "train_r2: 0.9996412018774883 test_r2: -2.461523170801259 independent_r2: 0.6724462000703733 runtime: 0.04538893699645996\n",
      "fold: 5\n",
      "train_r2: 0.9996491027567376 test_r2: -3.733548226380009 independent_r2: 0.6785319979325588 runtime: 0.03601360321044922\n",
      "fold: 6\n",
      "train_r2: 0.9996426805066739 test_r2: -2.4647066866658744 independent_r2: 0.6980703705274098 runtime: 0.04020571708679199\n",
      "fold: 7\n",
      "train_r2: 0.9996331318950937 test_r2: -2.089287475776321 independent_r2: 0.6260953528975055 runtime: 0.042127370834350586\n",
      "fold: 8\n",
      "train_r2: 0.9996331830511911 test_r2: -1.9650412017759606 independent_r2: 0.6740199272712479 runtime: 0.044768333435058594\n",
      "fold: 9\n",
      "train_r2: 0.9996319856606297 test_r2: -2.0203513444809413 independent_r2: 0.5063529951683914 runtime: 0.04714250564575195\n",
      "0.01 1\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9994551584163075 test_r2: -6.913106715603041 independent_r2: 0.3327203405843896 runtime: 0.026901960372924805\n",
      "fold: 1\n",
      "train_r2: 0.9994444239605987 test_r2: -5.4825999699600585 independent_r2: 0.31966235928238595 runtime: 0.02277064323425293\n",
      "fold: 2\n",
      "train_r2: 0.9994463608553522 test_r2: -5.603517139081462 independent_r2: 0.4354853728525111 runtime: 0.020267009735107422\n",
      "fold: 3\n",
      "train_r2: 0.9994446459995137 test_r2: -6.440645858040482 independent_r2: 0.5603567312891999 runtime: 0.022591114044189453\n",
      "fold: 4\n",
      "train_r2: 0.9994340489488145 test_r2: -4.460510859214695 independent_r2: 0.48328789344707357 runtime: 0.020133256912231445\n",
      "fold: 5\n",
      "train_r2: 0.9994553759383776 test_r2: -7.140509497348839 independent_r2: 0.44721090603863467 runtime: 0.018559932708740234\n",
      "fold: 6\n",
      "train_r2: 0.9994319356199305 test_r2: -4.33961792462741 independent_r2: 0.5346666483917863 runtime: 0.01875591278076172\n",
      "fold: 7\n",
      "train_r2: 0.9994276319108953 test_r2: -3.940206960301807 independent_r2: 0.4020862379499017 runtime: 0.03203916549682617\n",
      "fold: 8\n",
      "train_r2: 0.9994320067955851 test_r2: -4.202792102253508 independent_r2: 0.4280663491022859 runtime: 0.050797462463378906\n",
      "fold: 9\n",
      "train_r2: 0.9994069146613472 test_r2: -3.1462948140684075 independent_r2: 0.32225512168076276 runtime: 0.05735325813293457\n",
      "0.1 1e-06\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.037124074259726525 test_r2: -0.03480406268807523 independent_r2: 0.03578170042186102 runtime: 0.047278642654418945\n",
      "fold: 1\n",
      "train_r2: 0.03330519106236485 test_r2: -0.0026921804310198194 independent_r2: 0.0335848862409196 runtime: 0.040610551834106445\n",
      "fold: 2\n",
      "train_r2: 0.0354181097513564 test_r2: 0.03736979087190939 independent_r2: 0.03561288792962103 runtime: 0.02236175537109375\n",
      "fold: 3\n",
      "train_r2: 0.03678478966440435 test_r2: -0.09519427779908218 independent_r2: 0.035933150334439445 runtime: 0.01818990707397461\n",
      "fold: 4\n",
      "train_r2: 0.034680801720672894 test_r2: 0.010010047748338136 independent_r2: 0.03478198670436927 runtime: 0.025080442428588867\n",
      "fold: 5\n",
      "train_r2: 0.03547154583348289 test_r2: 0.03741305229553227 independent_r2: 0.03593742819644463 runtime: 0.02717137336730957\n",
      "fold: 6\n",
      "train_r2: 0.035158000855770144 test_r2: -0.017010122456580845 independent_r2: 0.03495428570937009 runtime: 0.020371437072753906\n",
      "fold: 7\n",
      "train_r2: 0.037083906669779454 test_r2: 0.024087340674255664 independent_r2: 0.035664476054287575 runtime: 0.02146601676940918\n",
      "fold: 8\n",
      "train_r2: 0.03353798284464249 test_r2: 0.03962855855769243 independent_r2: 0.03420816141467997 runtime: 0.023636579513549805\n",
      "fold: 9\n",
      "train_r2: 0.03576517606710594 test_r2: 0.015082789736072133 independent_r2: 0.03397391233675806 runtime: 0.042810916900634766\n",
      "0.1 1e-05\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.26649277733681676 test_r2: 0.14609376150415943 independent_r2: 0.2599420857252025 runtime: 0.054082393646240234\n",
      "fold: 1\n",
      "train_r2: 0.24135999124907614 test_r2: 0.2640650413706809 independent_r2: 0.2469232343057156 runtime: 0.04888343811035156\n",
      "fold: 2\n",
      "train_r2: 0.2535963048099378 test_r2: 0.27163157340237787 independent_r2: 0.2551584343778186 runtime: 0.022559165954589844\n",
      "fold: 3\n",
      "train_r2: 0.26125203015510157 test_r2: 0.14830733083484415 independent_r2: 0.2599070605548113 runtime: 0.02822566032409668\n",
      "fold: 4\n",
      "train_r2: 0.2500286278270818 test_r2: 0.25353451450794173 independent_r2: 0.2522505952721448 runtime: 0.02710866928100586\n",
      "fold: 5\n",
      "train_r2: 0.2555932507934151 test_r2: 0.2951745414207657 independent_r2: 0.2585364904825582 runtime: 0.03413677215576172\n",
      "fold: 6\n",
      "train_r2: 0.25159383972742344 test_r2: 0.23140557685818697 independent_r2: 0.2532009812778697 runtime: 0.018099069595336914\n",
      "fold: 7\n",
      "train_r2: 0.26348491126517737 test_r2: 0.18717351099237456 independent_r2: 0.2543731366468508 runtime: 0.021272659301757812\n",
      "fold: 8\n",
      "train_r2: 0.2449268483306295 test_r2: 0.2837199656437869 independent_r2: 0.24918850872657783 runtime: 0.03792405128479004\n",
      "fold: 9\n",
      "train_r2: 0.26001230217099136 test_r2: 0.17934438677050313 independent_r2: 0.24805401848787556 runtime: 0.05937933921813965\n",
      "0.1 0.0001\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.6498029392579378 test_r2: 0.3700008339098738 independent_r2: 0.6279420496464672 runtime: 0.045491933822631836\n",
      "fold: 1\n",
      "train_r2: 0.6077938764988925 test_r2: 0.7238933092236659 independent_r2: 0.6216145575236458 runtime: 0.04523301124572754\n",
      "fold: 2\n",
      "train_r2: 0.6221045656689381 test_r2: 0.6634930142321983 independent_r2: 0.6256505676302426 runtime: 0.04456067085266113\n",
      "fold: 3\n",
      "train_r2: 0.6279853203389498 test_r2: 0.5587848264309514 independent_r2: 0.6265806775824514 runtime: 0.049989938735961914\n",
      "fold: 4\n",
      "train_r2: 0.6205420419384933 test_r2: 0.6540023111032813 independent_r2: 0.6246617032493574 runtime: 0.0608057975769043\n",
      "fold: 5\n",
      "train_r2: 0.6327325609390264 test_r2: 0.6160652867045646 independent_r2: 0.6317290335876814 runtime: 0.03966164588928223\n",
      "fold: 6\n",
      "train_r2: 0.6164317325501598 test_r2: 0.6752846731431348 independent_r2: 0.6232792940408225 runtime: 0.0406641960144043\n",
      "fold: 7\n",
      "train_r2: 0.6337411549506544 test_r2: 0.5064675840918332 independent_r2: 0.6184080814077404 runtime: 0.0399327278137207\n",
      "fold: 8\n",
      "train_r2: 0.6220767758503001 test_r2: 0.6502533803728183 independent_r2: 0.6251719280712735 runtime: 0.043685197830200195\n",
      "fold: 9\n",
      "train_r2: 0.653030988598651 test_r2: 0.4421410364341607 independent_r2: 0.6191555887198793 runtime: 0.04551267623901367\n",
      "0.1 0.001\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.8249647661065789 test_r2: 0.3957648283689962 independent_r2: 0.789656772127088 runtime: 0.04641842842102051\n",
      "fold: 1\n",
      "train_r2: 0.8046152781391265 test_r2: 0.7945351349807142 independent_r2: 0.8043776968765691 runtime: 0.03939175605773926\n",
      "fold: 2\n",
      "train_r2: 0.8100587361724998 test_r2: 0.7452881951040875 independent_r2: 0.8045321959888381 runtime: 0.03025197982788086\n",
      "fold: 3\n",
      "train_r2: 0.8073711498064277 test_r2: 0.7193791525809345 independent_r2: 0.8035654969787356 runtime: 0.028293371200561523\n",
      "fold: 4\n",
      "train_r2: 0.8120455965496377 test_r2: 0.7175989204802591 independent_r2: 0.8035910232271175 runtime: 0.029155254364013672\n",
      "fold: 5\n",
      "train_r2: 0.8184996213853334 test_r2: 0.5859908295499221 independent_r2: 0.8027882652886683 runtime: 0.02819085121154785\n",
      "fold: 6\n",
      "train_r2: 0.8076239711982164 test_r2: 0.776680794996662 independent_r2: 0.8057951949978166 runtime: 0.029039382934570312\n",
      "fold: 7\n",
      "train_r2: 0.8150750135591891 test_r2: 0.6307443828332868 independent_r2: 0.7928134622676548 runtime: 0.022220849990844727\n",
      "fold: 8\n",
      "train_r2: 0.8115967436145411 test_r2: 0.6825011876367559 independent_r2: 0.7974182876542342 runtime: 0.019408464431762695\n",
      "fold: 9\n",
      "train_r2: 0.829529295514347 test_r2: 0.5075385656573846 independent_r2: 0.7772163732533917 runtime: 0.019424915313720703\n",
      "0.1 0.01\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9762974197786609 test_r2: 0.42936252138611086 independent_r2: 0.9303271086423022 runtime: 0.025577306747436523\n",
      "fold: 1\n",
      "train_r2: 0.9751407584295222 test_r2: 0.7325522446844865 independent_r2: 0.9498043092163169 runtime: 0.03758358955383301\n",
      "fold: 2\n",
      "train_r2: 0.975375979419272 test_r2: 0.7442924652777568 independent_r2: 0.9556397872235535 runtime: 0.07372498512268066\n",
      "fold: 3\n",
      "train_r2: 0.9742827479820589 test_r2: 0.7440348897728184 independent_r2: 0.9608794307733824 runtime: 0.047887563705444336\n",
      "fold: 4\n",
      "train_r2: 0.9760158485601944 test_r2: 0.6780735421291039 independent_r2: 0.9479107858322013 runtime: 0.04868316650390625\n",
      "fold: 5\n",
      "train_r2: 0.9767936330983483 test_r2: 0.49465399664692344 independent_r2: 0.9440915762119299 runtime: 0.020337343215942383\n",
      "fold: 6\n",
      "train_r2: 0.9749740951444482 test_r2: 0.7652696699747049 independent_r2: 0.9568316922141934 runtime: 0.025159358978271484\n",
      "fold: 7\n",
      "train_r2: 0.975194012638818 test_r2: 0.6676877915790653 independent_r2: 0.938011737573981 runtime: 0.03111410140991211\n",
      "fold: 8\n",
      "train_r2: 0.9749137322873649 test_r2: 0.6738655303905035 independent_r2: 0.9418493289740527 runtime: 0.022101163864135742\n",
      "fold: 9\n",
      "train_r2: 0.9774371700900358 test_r2: 0.49569474949855286 independent_r2: 0.8987872124876419 runtime: 0.026758193969726562\n",
      "0.1 0.1\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9697185435664422 test_r2: -4.240637727111256 independent_r2: 0.5308258384578118 runtime: 0.04602551460266113\n",
      "fold: 1\n",
      "train_r2: 0.9691229475361633 test_r2: -3.299265429838239 independent_r2: 0.5216184074523584 runtime: 0.057862043380737305\n",
      "fold: 2\n",
      "train_r2: 0.9691259694679143 test_r2: -3.089940873801062 independent_r2: 0.6224408064785278 runtime: 0.04276728630065918\n",
      "fold: 3\n",
      "train_r2: 0.9693979043901737 test_r2: -3.682832849985889 independent_r2: 0.6950575493076324 runtime: 0.04413747787475586\n",
      "fold: 4\n",
      "train_r2: 0.968841352213849 test_r2: -2.568679122492235 independent_r2: 0.6345057834272911 runtime: 0.03761696815490723\n",
      "fold: 5\n",
      "train_r2: 0.9695453140343759 test_r2: -3.9096015489340674 independent_r2: 0.6385368944882172 runtime: 0.06099820137023926\n",
      "fold: 6\n",
      "train_r2: 0.9689522384213617 test_r2: -2.5631209893246534 independent_r2: 0.6616224936410269 runtime: 0.042456626892089844\n",
      "fold: 7\n",
      "train_r2: 0.9681565545075814 test_r2: -2.1888584602301355 independent_r2: 0.5863892123528012 runtime: 0.046363115310668945\n",
      "fold: 8\n",
      "train_r2: 0.9681637873845317 test_r2: -2.078480184935842 independent_r2: 0.6335477674854827 runtime: 0.06001543998718262\n",
      "fold: 9\n",
      "train_r2: 0.9679865648629127 test_r2: -2.080605271470961 independent_r2: 0.4700867863994542 runtime: 0.09111428260803223\n",
      "0.1 1\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9540667025149729 test_r2: -6.913106715691002 independent_r2: 0.29137888305049475 runtime: 0.05651140213012695\n",
      "fold: 1\n",
      "train_r2: 0.95316172580205 test_r2: -5.482599971863453 independent_r2: 0.2784275491269971 runtime: 0.061278343200683594\n",
      "fold: 2\n",
      "train_r2: 0.9533250172349472 test_r2: -5.603517141417358 independent_r2: 0.39330460906185327 runtime: 0.05121445655822754\n",
      "fold: 3\n",
      "train_r2: 0.9531804449635778 test_r2: -6.44064585804053 independent_r2: 0.5171561141309347 runtime: 0.05284285545349121\n",
      "fold: 4\n",
      "train_r2: 0.95228705228426 test_r2: -4.460510859214856 independent_r2: 0.4407167155088958 runtime: 0.04620075225830078\n",
      "fold: 5\n",
      "train_r2: 0.9540850408838445 test_r2: -7.14050949736399 independent_r2: 0.4049343772916111 runtime: 0.05269575119018555\n",
      "fold: 6\n",
      "train_r2: 0.9521088864334901 test_r2: -4.339617924627554 independent_r2: 0.4916758481477913 runtime: 0.04865241050720215\n",
      "fold: 7\n",
      "train_r2: 0.9517460588642093 test_r2: -3.9402069603018095 independent_r2: 0.36017825292003025 runtime: 0.05164194107055664\n",
      "fold: 8\n",
      "train_r2: 0.9521148869530114 test_r2: -4.202792102330572 independent_r2: 0.38594617842038736 runtime: 0.04545283317565918\n",
      "fold: 9\n",
      "train_r2: 0.9499994748752462 test_r2: -3.146294814069628 independent_r2: 0.28099913604460924 runtime: 0.04304170608520508\n",
      "1 1e-06\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.003693262533276398 test_r2: -0.06903677730238877 independent_r2: 0.00244714360482301 runtime: 0.049530744552612305\n",
      "fold: 1\n",
      "train_r2: 0.0032890184217194696 test_r2: -0.04572390294247031 independent_r2: 0.0023296510417758487 runtime: 0.05685019493103027\n",
      "fold: 2\n",
      "train_r2: 0.0035141952929593767 test_r2: 0.0038990511325736055 independent_r2: 0.0035760793818342407 runtime: 0.09406423568725586\n",
      "fold: 3\n",
      "train_r2: 0.0036620935622928563 test_r2: -0.11938207018024327 independent_r2: 0.003576325359046839 runtime: 0.08771753311157227\n",
      "fold: 4\n",
      "train_r2: 0.0034311988971429397 test_r2: -0.021116205215712114 independent_r2: 0.0036228192354095246 runtime: 0.04432201385498047\n",
      "fold: 5\n",
      "train_r2: 0.0035220890663767435 test_r2: -0.0036958464731777774 independent_r2: 0.0033776307323891697 runtime: 0.050879716873168945\n",
      "fold: 6\n",
      "train_r2: 0.0034863544090617538 test_r2: -0.046697693448988 independent_r2: 0.0035977377649878406 runtime: 0.06382274627685547\n",
      "fold: 7\n",
      "train_r2: 0.0036868121029289824 test_r2: 0.00019575343744360207 independent_r2: 0.003422145628420692 runtime: 0.05160856246948242\n",
      "fold: 8\n",
      "train_r2: 0.0033071893386008577 test_r2: 0.003870368053052853 independent_r2: 0.0033703270978205913 runtime: 0.03844285011291504\n",
      "fold: 9\n",
      "train_r2: 0.003532313574478807 test_r2: -0.01061787658859159 independent_r2: 0.0028610588848206886 runtime: 0.044870615005493164\n",
      "1 1e-05\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.03691651442585442 test_r2: -0.04145522373814181 independent_r2: 0.03503219950606817 runtime: 0.04664278030395508\n",
      "fold: 1\n",
      "train_r2: 0.03310399287461929 test_r2: -0.007740842809927706 independent_r2: 0.032876169138022826 runtime: 0.05145907402038574\n",
      "fold: 2\n",
      "train_r2: 0.03520795097799678 test_r2: 0.03774332044957063 independent_r2: 0.035452588264470064 runtime: 0.07647538185119629\n",
      "fold: 3\n",
      "train_r2: 0.03659466931694355 test_r2: -0.08417707729269464 independent_r2: 0.03640582188726449 runtime: 0.046892642974853516\n",
      "fold: 4\n",
      "train_r2: 0.03448174306114182 test_r2: 0.013987005920325979 independent_r2: 0.03497820323164591 runtime: 0.05028200149536133\n",
      "fold: 5\n",
      "train_r2: 0.03526973599083494 test_r2: 0.035138242967170163 independent_r2: 0.03559504848839368 runtime: 0.051442623138427734\n",
      "fold: 6\n",
      "train_r2: 0.03496520676456738 test_r2: -0.011394797114978816 independent_r2: 0.035267959004744664 runtime: 0.05033707618713379\n",
      "fold: 7\n",
      "train_r2: 0.036870066161434734 test_r2: 0.023186632728599 independent_r2: 0.03536760765719171 runtime: 0.046983957290649414\n",
      "fold: 8\n",
      "train_r2: 0.033336401219711154 test_r2: 0.03935778097682974 independent_r2: 0.03399898018582692 runtime: 0.048648834228515625\n",
      "fold: 9\n",
      "train_r2: 0.03556242690130007 test_r2: 0.013013336246791463 independent_r2: 0.03346659214717018 runtime: 0.06749415397644043\n",
      "1 0.0001\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.2649837799130488 test_r2: 0.13925545453390387 independent_r2: 0.2579914185802179 runtime: 0.05616259574890137\n",
      "fold: 1\n",
      "train_r2: 0.24004244956853238 test_r2: 0.2573944309757401 independent_r2: 0.24504983250737788 runtime: 0.04463076591491699\n",
      "fold: 2\n",
      "train_r2: 0.25216988983797795 test_r2: 0.2714838480365457 independent_r2: 0.2538412741825893 runtime: 0.04818439483642578\n",
      "fold: 3\n",
      "train_r2: 0.25993675046257614 test_r2: 0.15667797394708693 independent_r2: 0.25917288813704753 runtime: 0.04726243019104004\n",
      "fold: 4\n",
      "train_r2: 0.24869471610682925 test_r2: 0.25587166518978866 independent_r2: 0.25126708440195433 runtime: 0.05951642990112305\n",
      "fold: 5\n",
      "train_r2: 0.2541652177051993 test_r2: 0.2922645501253981 independent_r2: 0.25700841073544634 runtime: 0.05355429649353027\n",
      "fold: 6\n",
      "train_r2: 0.2502904898668078 test_r2: 0.23494859579329164 independent_r2: 0.2523253703984393 runtime: 0.04574227333068848\n",
      "fold: 7\n",
      "train_r2: 0.2620092399229108 test_r2: 0.1857955973324179 independent_r2: 0.2529095202524614 runtime: 0.04507780075073242\n",
      "fold: 8\n",
      "train_r2: 0.24357023042889547 test_r2: 0.28205499418808977 independent_r2: 0.24779802578151844 runtime: 0.04693245887756348\n",
      "fold: 9\n",
      "train_r2: 0.25866018538087165 test_r2: 0.17588004586917538 independent_r2: 0.24635911949981437 runtime: 0.0510706901550293\n",
      "1 0.001\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.6451980276564442 test_r2: 0.3695222389764682 independent_r2: 0.6237074006435348 runtime: 0.04102897644042969\n",
      "fold: 1\n",
      "train_r2: 0.6038812543346376 test_r2: 0.7095098379186029 independent_r2: 0.6166202489024237 runtime: 0.06523728370666504\n",
      "fold: 2\n",
      "train_r2: 0.6179718707607816 test_r2: 0.6622236423463326 independent_r2: 0.6217625502882478 runtime: 0.0466158390045166\n",
      "fold: 3\n",
      "train_r2: 0.6243921322219417 test_r2: 0.5588237158088127 independent_r2: 0.6232277230697926 runtime: 0.04838395118713379\n",
      "fold: 4\n",
      "train_r2: 0.6163330174092814 test_r2: 0.653076238221705 independent_r2: 0.6207736377562268 runtime: 0.0725867748260498\n",
      "fold: 5\n",
      "train_r2: 0.6285075855053999 test_r2: 0.6162927166890673 independent_r2: 0.627807589195992 runtime: 0.054628610610961914\n",
      "fold: 6\n",
      "train_r2: 0.6123848817891113 test_r2: 0.6733566629674428 independent_r2: 0.6194350828302264 runtime: 0.04688072204589844\n",
      "fold: 7\n",
      "train_r2: 0.6294733088119451 test_r2: 0.5038819939907884 independent_r2: 0.6143443422661582 runtime: 0.05458354949951172\n",
      "fold: 8\n",
      "train_r2: 0.617764047224017 test_r2: 0.6467180160001302 independent_r2: 0.6209445837182768 runtime: 0.03907895088195801\n",
      "fold: 9\n",
      "train_r2: 0.6489349632608952 test_r2: 0.4336909454856982 independent_r2: 0.6143551181269393 runtime: 0.02399587631225586\n",
      "1 0.01\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.8147684672333761 test_r2: 0.41703301672962334 independent_r2: 0.7821617368379665 runtime: 0.02582240104675293\n",
      "fold: 1\n",
      "train_r2: 0.7954794684174079 test_r2: 0.7582455608742533 independent_r2: 0.7924325401617112 runtime: 0.02703094482421875\n",
      "fold: 2\n",
      "train_r2: 0.8010819686055164 test_r2: 0.7513367312195932 independent_r2: 0.7968390052791783 runtime: 0.021201372146606445\n",
      "fold: 3\n",
      "train_r2: 0.8023211259095181 test_r2: 0.6672818930601051 independent_r2: 0.7957752576060403 runtime: 0.024063825607299805\n",
      "fold: 4\n",
      "train_r2: 0.8023314082597722 test_r2: 0.7220838863259162 independent_r2: 0.7952436180955547 runtime: 0.04411435127258301\n",
      "fold: 5\n",
      "train_r2: 0.8107484163633932 test_r2: 0.5879386279216434 independent_r2: 0.7956977611638016 runtime: 0.046751976013183594\n",
      "fold: 6\n",
      "train_r2: 0.7977655892517151 test_r2: 0.7816677648865922 independent_r2: 0.7972734302556292 runtime: 0.04640793800354004\n",
      "fold: 7\n",
      "train_r2: 0.8064263040032746 test_r2: 0.6191364610370447 independent_r2: 0.7838082675067405 runtime: 0.050750732421875\n",
      "fold: 8\n",
      "train_r2: 0.8024544787396568 test_r2: 0.6794516269146291 independent_r2: 0.7889452025534742 runtime: 0.10987687110900879\n",
      "fold: 9\n",
      "train_r2: 0.8245506827539323 test_r2: 0.44983459282154314 independent_r2: 0.763633872360681 runtime: 0.04862213134765625\n",
      "1 0.1\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: -0.07974517328598552 test_r2: -5.189122682575396 independent_r2: -0.5049862516195303 runtime: 0.06231260299682617\n",
      "fold: 1\n",
      "train_r2: -0.10159819025691186 test_r2: -4.064222105721793 independent_r2: -0.5125454787487407 runtime: 0.07079267501831055\n",
      "fold: 2\n",
      "train_r2: -0.09938694385171165 test_r2: -3.9686097112018377 independent_r2: -0.42982635936567504 runtime: 0.08578252792358398\n",
      "fold: 3\n",
      "train_r2: -0.0943351737333491 test_r2: -4.647254656345804 independent_r2: -0.35515314282699073 runtime: 0.0684671401977539\n",
      "fold: 4\n",
      "train_r2: -0.11483901135625385 test_r2: -3.2308017889683214 independent_r2: -0.406591500329774 runtime: 0.0739438533782959\n",
      "fold: 5\n",
      "train_r2: -0.08473664265639136 test_r2: -5.0178023050297975 independent_r2: -0.41903784968331736 runtime: 0.07020854949951172\n",
      "fold: 6\n",
      "train_r2: -0.11439085623973266 test_r2: -3.176066072427882 independent_r2: -0.3759018517229371 runtime: 0.05363750457763672\n",
      "fold: 7\n",
      "train_r2: -0.13539544524333702 test_r2: -2.803926179060116 independent_r2: -0.45791677255817365 runtime: 0.04838824272155762\n",
      "fold: 8\n",
      "train_r2: -0.1334194355866305 test_r2: -2.7998262108440968 independent_r2: -0.4262721922640611 runtime: 0.05231046676635742\n",
      "fold: 9\n",
      "train_r2: -0.1555909645095459 test_r2: -2.4526372613510596 independent_r2: -0.5288851921097857 runtime: 0.023254871368408203\n",
      "1 1\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: -0.3894822494785084 test_r2: -6.913106716135217 independent_r2: -0.9323744002960435 runtime: 0.03593039512634277\n",
      "fold: 1\n",
      "train_r2: -0.41685779451213967 test_r2: -5.482599981475601 independent_r2: -0.9421688475867411 runtime: 0.06320714950561523\n",
      "fold: 2\n",
      "train_r2: -0.4119182286666574 test_r2: -5.603517153213627 independent_r2: -0.8552930710104667 runtime: 0.0690457820892334\n",
      "fold: 3\n",
      "train_r2: -0.4162915404145764 test_r2: -6.440645858040771 independent_r2: -0.7616303692524817 runtime: 0.05389142036437988\n",
      "fold: 4\n",
      "train_r2: -0.4433166689831225 test_r2: -4.460510859215665 independent_r2: -0.819437664460464 runtime: 0.07610058784484863\n",
      "fold: 5\n",
      "train_r2: -0.3889275138249093 test_r2: -7.140509497440487 independent_r2: -0.8464980577426244 runtime: 0.07644104957580566\n",
      "fold: 6\n",
      "train_r2: -0.44870618596538514 test_r2: -4.339617924628285 independent_r2: -0.7808998204022926 runtime: 0.07434368133544922\n",
      "fold: 7\n",
      "train_r2: -0.45968171995561224 test_r2: -3.940206960301821 independent_r2: -0.8803448767932149 runtime: 0.07608914375305176\n",
      "fold: 8\n",
      "train_r2: -0.4485246702377812 test_r2: -4.202792102719742 independent_r2: -0.8608578831780926 runtime: 0.0531458854675293\n",
      "fold: 9\n",
      "train_r2: -0.512515885652582 test_r2: -3.1462948140757936 independent_r2: -0.9402240839312996 runtime: 0.04398179054260254\n",
      "10 1e-06\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: -0.014331542150836096 test_r2: -0.1513950862739879 independent_r2: -0.020910276086356205 runtime: 0.0466005802154541\n",
      "fold: 1\n",
      "train_r2: -0.014724586513631843 test_r2: -0.10990289325517888 independent_r2: -0.020449847970999135 runtime: 0.04191780090332031\n",
      "fold: 2\n",
      "train_r2: -0.014639054570162235 test_r2: -0.013081096490959565 independent_r2: -0.0144764469372749 runtime: 0.061094045639038086\n",
      "fold: 3\n",
      "train_r2: -0.014678316706683958 test_r2: -0.04107170235664048 independent_r2: -0.008927951475826701 runtime: 0.04154181480407715\n",
      "fold: 4\n",
      "train_r2: -0.015050873644501861 test_r2: -0.0014112352342121603 independent_r2: -0.011202709358354879 runtime: 0.05240678787231445\n",
      "fold: 5\n",
      "train_r2: -0.014175793805166093 test_r2: -0.04879742362301087 independent_r2: -0.0161732863955808 runtime: 0.04525327682495117\n",
      "fold: 6\n",
      "train_r2: -0.014940031857927805 test_r2: -0.011258032323801181 independent_r2: -0.010056735381351878 runtime: 0.04064464569091797\n",
      "fold: 7\n",
      "train_r2: -0.015061782347278063 test_r2: -0.021649411130000296 independent_r2: -0.015697947847027383 runtime: 0.05731964111328125\n",
      "fold: 8\n",
      "train_r2: -0.014956839840911229 test_r2: -0.013805726586295197 independent_r2: -0.014829105107374563 runtime: 0.039040565490722656\n",
      "fold: 9\n",
      "train_r2: -0.015749495986283835 test_r2: -0.03969327488940211 independent_r2: -0.017988686349372562 runtime: 0.04032182693481445\n",
      "10 1e-05\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: -0.010903736020848598 test_r2: -0.14850820069799564 independent_r2: -0.01754484740106199 runtime: 0.04809880256652832\n",
      "fold: 1\n",
      "train_r2: -0.011650317517152065 test_r2: -0.10601549838061053 independent_r2: -0.017303199951860915 runtime: 0.06555914878845215\n",
      "fold: 2\n",
      "train_r2: -0.011370685924415591 test_r2: -0.009485508189397418 independent_r2: -0.011180225528021115 runtime: 0.054381370544433594\n",
      "fold: 3\n",
      "train_r2: -0.01125629810912887 test_r2: -0.037511309980634344 independent_r2: -0.00552241289725619 runtime: 0.04743194580078125\n",
      "fold: 4\n",
      "train_r2: -0.011839143654420958 test_r2: 0.002197898471456128 independent_r2: -0.007961507610104945 runtime: 0.05107522010803223\n",
      "fold: 5\n",
      "train_r2: -0.010901170833539808 test_r2: -0.04470355780051882 independent_r2: -0.012844217527975044 runtime: 0.051186323165893555\n",
      "fold: 6\n",
      "train_r2: -0.01167475802938811 test_r2: -0.007662531690367791 independent_r2: -0.0067773944920501705 runtime: 0.053566932678222656\n",
      "fold: 7\n",
      "train_r2: -0.011632311448250832 test_r2: -0.01923296800352814 independent_r2: -0.01239152279249578 runtime: 0.07366561889648438\n",
      "fold: 8\n",
      "train_r2: -0.011861768938954409 test_r2: -0.01014867480700099 independent_r2: -0.011672315248939302 runtime: 0.04945492744445801\n",
      "fold: 9\n",
      "train_r2: -0.012436421932128017 test_r2: -0.03731928125868822 independent_r2: -0.01483445305475084 runtime: 0.04924273490905762\n",
      "10 0.0001\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.02169162788053014 test_r2: -0.12120632574803514 independent_r2: 0.014444738661558376 runtime: 0.05449652671813965\n",
      "fold: 1\n",
      "train_r2: 0.017625876102694837 test_r2: -0.06900451097906912 independent_r2: 0.012661357081265479 runtime: 0.05324387550354004\n",
      "fold: 2\n",
      "train_r2: 0.01969623639383522 test_r2: 0.02474726925858406 independent_r2: 0.020156188846753964 runtime: 0.05135083198547363\n",
      "fold: 3\n",
      "train_r2: 0.02123074192420049 test_r2: -0.003666772010475494 independent_r2: 0.026810760996697103 runtime: 0.05871868133544922\n",
      "fold: 4\n",
      "train_r2: 0.01872000159245235 test_r2: 0.036531565697863755 independent_r2: 0.022877422184300222 runtime: 0.05835556983947754\n",
      "fold: 5\n",
      "train_r2: 0.020254303947087715 test_r2: -0.00590985939489852 independent_r2: 0.018818668070059164 runtime: 0.04940319061279297\n",
      "fold: 6\n",
      "train_r2: 0.019356366144457682 test_r2: 0.026598316945205647 independent_r2: 0.024395375468777236 runtime: 0.05018782615661621\n",
      "fold: 7\n",
      "train_r2: 0.020925031208783418 test_r2: 0.003838014871865525 independent_r2: 0.019013503954911526 runtime: 0.055690765380859375\n",
      "fold: 8\n",
      "train_r2: 0.01764488061947922 test_r2: 0.024613724319348607 independent_r2: 0.01841154073417961 runtime: 0.0633535385131836\n",
      "fold: 9\n",
      "train_r2: 0.019129063993582274 test_r2: -0.014759677767658408 independent_r2: 0.015208073974199543 runtime: 0.04562115669250488\n",
      "10 0.001\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.23444884473630523 test_r2: 0.04861495851361464 independent_r2: 0.22254143346203092 runtime: 0.05298328399658203\n",
      "fold: 1\n",
      "train_r2: 0.210999743608004 test_r2: 0.1739828805993996 independent_r2: 0.2104271816719544 runtime: 0.0487978458404541\n",
      "fold: 2\n",
      "train_r2: 0.22207506782696895 test_r2: 0.25069110014378604 independent_r2: 0.22454182122045363 runtime: 0.053559303283691406\n",
      "fold: 3\n",
      "train_r2: 0.23094762892952025 test_r2: 0.21746209419954798 independent_r2: 0.23569070367155653 runtime: 0.03889584541320801\n",
      "fold: 4\n",
      "train_r2: 0.2191901124354021 test_r2: 0.2613057279266201 independent_r2: 0.22513973021710398 runtime: 0.024453401565551758\n",
      "fold: 5\n",
      "train_r2: 0.22457290099466753 test_r2: 0.2395165493237319 independent_r2: 0.22585538100368518 runtime: 0.02755594253540039\n",
      "fold: 6\n",
      "train_r2: 0.22118024929139557 test_r2: 0.2545107901146739 independent_r2: 0.2275829659791142 runtime: 0.02418375015258789\n",
      "fold: 7\n",
      "train_r2: 0.23098825289247438 test_r2: 0.16000947197677318 independent_r2: 0.2225264799312031 runtime: 0.029253482818603516\n",
      "fold: 8\n",
      "train_r2: 0.21396205673802482 test_r2: 0.24989645988817777 independent_r2: 0.21790978146622475 runtime: 0.0408322811126709\n",
      "fold: 9\n",
      "train_r2: 0.22824423604713973 test_r2: 0.13080487371414284 independent_r2: 0.2135988139421633 runtime: 0.05899238586425781\n",
      "10 0.01\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.4961464960542816 test_r2: 0.14606192776874016 independent_r2: 0.4691163684611386 runtime: 0.04681229591369629\n",
      "fold: 1\n",
      "train_r2: 0.4610466842563796 test_r2: 0.4062413651505601 independent_r2: 0.4575597891962171 runtime: 0.033418893814086914\n",
      "fold: 2\n",
      "train_r2: 0.47157048429004766 test_r2: 0.5540120223627478 independent_r2: 0.4786272253012528 runtime: 0.024371623992919922\n",
      "fold: 3\n",
      "train_r2: 0.4859363205338172 test_r2: 0.5333940004892022 independent_r2: 0.49243957508851377 runtime: 0.02415299415588379\n",
      "fold: 4\n",
      "train_r2: 0.4711429944586867 test_r2: 0.5336888842179471 independent_r2: 0.47838882894169055 runtime: 0.026162385940551758\n",
      "fold: 5\n",
      "train_r2: 0.48286207274742365 test_r2: 0.4691758345641418 independent_r2: 0.48211271270780354 runtime: 0.023768901824951172\n",
      "fold: 6\n",
      "train_r2: 0.46939309943937435 test_r2: 0.5814423219179066 independent_r2: 0.4815324611998528 runtime: 0.05743527412414551\n",
      "fold: 7\n",
      "train_r2: 0.48119869172552887 test_r2: 0.39529119150669734 independent_r2: 0.470892060229391 runtime: 0.0517878532409668\n",
      "fold: 8\n",
      "train_r2: 0.4690258009565963 test_r2: 0.483978662289953 independent_r2: 0.4706687728315595 runtime: 0.025585412979125977\n",
      "fold: 9\n",
      "train_r2: 0.4987003234385933 test_r2: 0.23978083435319064 independent_r2: 0.457233852297 runtime: 0.02433323860168457\n",
      "10 0.1\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: -3.3646483117540713 test_r2: -6.535617148128663 independent_r2: -3.6104554647390357 runtime: 0.02581167221069336\n",
      "fold: 1\n",
      "train_r2: -3.4518900325554833 test_r2: -5.169016933247242 independent_r2: -3.613294930285683 runtime: 0.027496814727783203\n",
      "fold: 2\n",
      "train_r2: -3.4370415356036386 test_r2: -5.241717915953306 independent_r2: -3.591050274410705 runtime: 0.027652502059936523\n",
      "fold: 3\n",
      "train_r2: -3.443485611865616 test_r2: -6.042973662839984 independent_r2: -3.5648962864713756 runtime: 0.030434608459472656\n",
      "fold: 4\n",
      "train_r2: -3.5280747111405715 test_r2: -4.188228280071599 independent_r2: -3.5790668570393613 runtime: 0.05115866661071777\n",
      "fold: 5\n",
      "train_r2: -3.3689657008567044 test_r2: -6.665270157000116 independent_r2: -3.5910852774101594 runtime: 0.0484926700592041\n",
      "fold: 6\n",
      "train_r2: -3.5417685999007116 test_r2: -4.079040560431648 independent_r2: -3.5681204547199377 runtime: 0.05450558662414551\n",
      "fold: 7\n",
      "train_r2: -3.586523390898421 test_r2: -3.6895825815656274 independent_r2: -3.5982611047328223 runtime: 0.05248832702636719\n",
      "fold: 8\n",
      "train_r2: -3.5586350424493762 test_r2: -3.8870978241467933 independent_r2: -3.594704580420168 runtime: 0.04935789108276367\n",
      "fold: 9\n",
      "train_r2: -3.735954256825485 test_r2: -2.992159765460846 independent_r2: -3.606669392805335 runtime: 0.04952049255371094\n",
      "10 1\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: -3.5933297521816536 test_r2: -6.91310671657943 independent_r2: -3.8505553083214714 runtime: 0.052831172943115234\n",
      "fold: 1\n",
      "train_r2: -3.683827419954704 test_r2: -5.48259999108775 independent_r2: -3.8528217922701264 runtime: 0.05363011360168457\n",
      "fold: 2\n",
      "train_r2: -3.6674982766627204 test_r2: -5.603517165009898 independent_r2: -3.8327183067812527 runtime: 0.04812765121459961\n",
      "fold: 3\n",
      "train_r2: -3.6819555073632513 test_r2: -6.440645858041012 independent_r2: -3.8110442928626522 runtime: 0.057619571685791016\n",
      "fold: 4\n",
      "train_r2: -3.7712947754218655 test_r2: -4.460510859216475 independent_r2: -3.8244211876215415 runtime: 0.05034184455871582\n",
      "fold: 5\n",
      "train_r2: -3.5914959153260115 test_r2: -7.140509497516989 independent_r2: -3.8306830968122725 runtime: 0.05699896812438965\n",
      "fold: 6\n",
      "train_r2: -3.789111360475535 test_r2: -4.339617924629013 independent_r2: -3.815503339409733 runtime: 0.038027048110961914\n",
      "fold: 7\n",
      "train_r2: -3.825394117532432 test_r2: -3.9402069603018317 independent_r2: -3.838515418574553 runtime: 0.024401187896728516\n",
      "fold: 8\n",
      "train_r2: -3.788511308443513 test_r2: -4.20279210310891 independent_r2: -3.83400603163454 runtime: 0.02471184730529785\n",
      "fold: 9\n",
      "train_r2: -4.000052516632604 test_r2: -3.1462948140819584 independent_r2: -3.852371764028283 runtime: 0.025107383728027344\n"
     ]
    }
   ],
   "source": [
    "import itertools\n",
    "feature_sets = ['RelativeExpr']\n",
    "methods = ['KernelRidge']\n",
    "targets = ['auc']\n",
    "# TODO: rbf kernel bandwith? median\n",
    "alpha_vals = [1e-3, 1e-2, 1e-1, 1, 10]\n",
    "gam_vals = [1e-6, 1e-5, 1e-4, 1e-3, 1e-2, 1e-1, 1]\n",
    "results_vals_rel = {}\n",
    "for dataset, method, target, alpha_val, gam_val in itertools.product(feature_sets, methods, targets, alpha_vals, gam_vals):\n",
    "    print(alpha_val, gam_val)\n",
    "    result_all = run_regression_dataset(dataset, dic_dataset_train, dic_dataset_train, method=method, target=target,\n",
    "                                       model_kwargs={'kernel': 'rbf', 'alpha':alpha_val, 'gamma': gam_val})\n",
    "    results_vals_rel[alpha_val, gam_val] = result_all"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 275,
   "metadata": {},
   "outputs": [],
   "source": [
    "mean_test_r2_rel = {k: np.mean([x['test_r2'] for iter, x in row.items()]) for k, row in results_vals_rel.items()}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 280,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[((0.001, 1e-05), 0.6561645415743705),\n",
       " ((0.01, 0.0001), 0.6561212853376683),\n",
       " ((0.1, 0.001), 0.6556021992189003),\n",
       " ((1, 0.01), 0.6434010161790943),\n",
       " ((0.1, 0.01), 0.6425487401340026),\n",
       " ((0.01, 0.01), 0.6276400754953211),\n",
       " ((0.001, 0.01), 0.6247678347984345),\n",
       " ((0.01, 0.001), 0.6141797648253172),\n",
       " ((0.001, 0.0001), 0.6019722051077785),\n",
       " ((0.001, 1e-06), 0.5863325459082701),\n",
       " ((0.01, 1e-05), 0.5863063916657675),\n",
       " ((0.1, 0.0001), 0.5860386255646483),\n",
       " ((1, 0.001), 0.5827096008405048),\n",
       " ((0.001, 0.001), 0.5028904334630672),\n",
       " ((10, 0.01), 0.4343067044621087),\n",
       " ((0.01, 1e-06), 0.22611609173710462),\n",
       " ((0.1, 1e-05), 0.22604502033056212),\n",
       " ((1, 0.0001), 0.2251627155991438),\n",
       " ((10, 0.001), 0.19867949064004678),\n",
       " ((1, 1e-05), 0.0017658378333544),\n",
       " ((0.1, 1e-06), 0.001389093650904194),\n",
       " ((10, 0.0001), -0.009821825480726909),\n",
       " ((1, 1e-06), -0.030830519952850178),\n",
       " ((10, 1e-05), -0.04183896323372858),\n",
       " ((10, 1e-06), -0.04520658821634886),\n",
       " ((0.001, 0.1), -2.834001606685836),\n",
       " ((0.01, 0.1), -2.847063720516568),\n",
       " ((0.1, 0.1), -2.9702022458124344),\n",
       " ((1, 0.1), -3.73502689735261),\n",
       " ((10, 0.1), -4.849070482884582),\n",
       " ((0.001, 1), -5.166980184001387),\n",
       " ((0.01, 1), -5.1669801840499705),\n",
       " ((0.1, 1), -5.1669801844920755),\n",
       " ((1, 1), -5.166980186724701),\n",
       " ((10, 1), -5.166980188957328)]"
      ]
     },
     "execution_count": 280,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted(list(mean_test_r2_rel.items()), key=lambda x: x[1])[::-1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For Kernel Ridge regression on gene expression, we use 0.1, 1e-5 as the values for alpha and gamma."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Run Venetoclax"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [],
   "source": [
    "import itertools"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 288,
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_sets = ['Mut', 'VAF', 'Mut+VAF', 'Expr', 'RelativeExpr', 'Mut+RelativeExpr', 'Mut+VAF+RelativeExpr']\n",
    "methods = ['XGBoost', 'LassoCV', 'KernelRidge']\n",
    "targets = ['auc', 'ic50']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mut XGBoost auc\n",
      "Mut\n",
      "fold: 0\n",
      "train_r2: 0.6494070356007863 test_r2: -0.26479927109415 independent_r2: -0.38114076250799767 runtime: 0.10109567642211914\n",
      "fold: 1\n",
      "train_r2: 0.6403535103931495 test_r2: -0.2469005673667326 independent_r2: -0.3018170082455991 runtime: 0.04544472694396973\n",
      "fold: 2\n",
      "train_r2: 0.6357409150853082 test_r2: -0.5127740468563029 independent_r2: -0.32358474249965274 runtime: 0.08223605155944824\n",
      "fold: 3\n",
      "train_r2: 0.6526175701589028 test_r2: -0.6437419685935695 independent_r2: -0.4101011711722087 runtime: 0.056340932846069336\n",
      "fold: 4\n",
      "train_r2: 0.6467427534228212 test_r2: -0.5601878667739335 independent_r2: -0.31946353075285794 runtime: 0.05835890769958496\n",
      "fold: 5\n",
      "train_r2: 0.6519738550478256 test_r2: -0.31807796947482414 independent_r2: -0.29297200786564304 runtime: 0.06537580490112305\n",
      "fold: 6\n",
      "train_r2: 0.6407780864214117 test_r2: -0.33945264666702224 independent_r2: -0.3063785952771443 runtime: 0.08284926414489746\n",
      "fold: 7\n",
      "train_r2: 0.668997935505319 test_r2: -0.07976822972803466 independent_r2: -0.29228150964789545 runtime: 0.05369973182678223\n",
      "fold: 8\n",
      "train_r2: 0.6445468962657712 test_r2: -0.3753597301944065 independent_r2: -0.3502462886230202 runtime: 0.053926944732666016\n",
      "fold: 9\n",
      "train_r2: 0.64943471271549 test_r2: 0.14334752934396955 independent_r2: -0.33417024229461134 runtime: 0.18523454666137695\n",
      "Mut XGBoost ic50\n",
      "Mut\n",
      "fold: 0\n",
      "train_r2: 0.6901673967079103 test_r2: -0.45130352771489934 independent_r2: -0.6106970092926327 runtime: 0.05608558654785156\n",
      "fold: 1\n",
      "train_r2: 0.6821814325471043 test_r2: -0.3789639800464424 independent_r2: -0.6090402100291787 runtime: 0.055796146392822266\n",
      "fold: 2\n",
      "train_r2: 0.6711823305174037 test_r2: -0.7038388626894339 independent_r2: -0.6060422098476519 runtime: 0.04863762855529785\n",
      "fold: 3\n",
      "train_r2: 0.6579771780415897 test_r2: -0.3813284747406289 independent_r2: -0.7463420927421174 runtime: 0.06187915802001953\n",
      "fold: 4\n",
      "train_r2: 0.6873935032212355 test_r2: -0.4655584015967813 independent_r2: -0.5065028435168204 runtime: 0.048673152923583984\n",
      "fold: 5\n",
      "train_r2: 0.676128670935879 test_r2: -0.17892680717684972 independent_r2: -0.5215685738456783 runtime: 0.06263566017150879\n",
      "fold: 6\n",
      "train_r2: 0.6883538902880486 test_r2: -0.5522200107556039 independent_r2: -0.5847194239685123 runtime: 0.05813765525817871\n",
      "fold: 7\n",
      "train_r2: 0.7263132457426058 test_r2: -0.8682361482375438 independent_r2: -0.6415866130887702 runtime: 0.10660672187805176\n",
      "fold: 8\n",
      "train_r2: 0.6856398864975991 test_r2: -0.10673005834371208 independent_r2: -0.568308131930451 runtime: 0.058518171310424805\n",
      "fold: 9\n",
      "train_r2: 0.6735067002857538 test_r2: -0.49242764442168974 independent_r2: -0.6853093156104169 runtime: 0.04718637466430664\n",
      "Mut LassoCV auc\n",
      "Mut\n",
      "fold: 0\n",
      "train_r2: 0.29536925705553074 test_r2: -0.25802867372986094 independent_r2: -0.005977743262371105 runtime: 0.05629897117614746\n",
      "fold: 1\n",
      "train_r2: 0.253476281951277 test_r2: 0.09136145071251234 independent_r2: 0.025128120051992053 runtime: 0.048125267028808594\n",
      "fold: 2\n",
      "train_r2: 0.2569531169982904 test_r2: 0.16089280734649847 independent_r2: -0.03223015367550364 runtime: 0.045662879943847656\n",
      "fold: 3\n",
      "train_r2: 0.2606190653131032 test_r2: 0.15348898641825715 independent_r2: -0.007264368225024853 runtime: 0.050795555114746094\n",
      "fold: 4\n",
      "train_r2: 0.26323549897259213 test_r2: 0.2910732713450137 independent_r2: -0.010901305624496205 runtime: 0.04576873779296875\n",
      "fold: 5\n",
      "train_r2: 0.27882151205616135 test_r2: 0.03315544894961808 independent_r2: -0.015069999424244251 runtime: 0.05031538009643555\n",
      "fold: 6\n",
      "train_r2: 0.25644341509878155 test_r2: 0.1031209152485707 independent_r2: -0.051747923028118814 runtime: 0.05121564865112305\n",
      "fold: 7\n",
      "train_r2: 0.26739699183554155 test_r2: 0.17793308214615078 independent_r2: 0.012116479321742313 runtime: 0.046140432357788086\n",
      "fold: 8\n",
      "train_r2: 0.3035728811155164 test_r2: -0.12058418174211472 independent_r2: -0.05343567439295227 runtime: 0.04600024223327637\n",
      "fold: 9\n",
      "train_r2: 0.22463701378511036 test_r2: 0.1648522047609522 independent_r2: -0.037902517214887244 runtime: 0.04398226737976074\n",
      "Mut LassoCV ic50\n",
      "Mut\n",
      "fold: 0\n",
      "train_r2: 0.251822517240576 test_r2: -0.22306739257226305 independent_r2: 0.03932105018653731 runtime: 0.04501986503601074\n",
      "fold: 1\n",
      "train_r2: 0.20167081049418367 test_r2: 0.0007951275362293453 independent_r2: 0.07363582176785699 runtime: 0.050753116607666016\n",
      "fold: 2\n",
      "train_r2: 0.2095106641500225 test_r2: 0.05842960429241717 independent_r2: 0.0029654612807470393 runtime: 0.04562854766845703\n",
      "fold: 3\n",
      "train_r2: 0.18893804255411162 test_r2: 0.15666975611516964 independent_r2: 0.03720790377908878 runtime: 0.054219961166381836\n",
      "fold: 4\n",
      "train_r2: 0.22969752988773762 test_r2: 0.044668226672653355 independent_r2: 0.04044206285101026 runtime: 0.05563187599182129\n",
      "fold: 5\n",
      "train_r2: 0.22056200051871233 test_r2: 0.15139440315674535 independent_r2: 0.009665842795093638 runtime: 0.0486910343170166\n",
      "fold: 6\n",
      "train_r2: 0.2217005034594276 test_r2: -0.2525763567134496 independent_r2: 0.00017531409125670638 runtime: 0.05505037307739258\n",
      "fold: 7\n",
      "train_r2: 0.23166801610131127 test_r2: 0.07193304467834816 independent_r2: 0.03448959468772406 runtime: 0.05110645294189453\n",
      "fold: 8\n",
      "train_r2: 0.21887154504779927 test_r2: 0.10515922917181009 independent_r2: 0.008842425442323831 runtime: 0.04736447334289551\n",
      "fold: 9\n",
      "train_r2: 0.1657639405648057 test_r2: 0.06970545318229515 independent_r2: 0.0015993655344370694 runtime: 0.04745650291442871\n",
      "Mut KernelRidge auc\n",
      "Mut\n",
      "fold: 0\n",
      "train_r2: 0.23762356116419803 test_r2: -0.049779021342046814 independent_r2: 0.013206143001593396 runtime: 0.03474116325378418\n",
      "fold: 1\n",
      "train_r2: 0.21751914082003743 test_r2: 0.05457285972220083 independent_r2: 0.031086655663216445 runtime: 0.010384559631347656\n",
      "fold: 2\n",
      "train_r2: 0.21747329151139283 test_r2: 0.11251459576278788 independent_r2: -0.010794543799918621 runtime: 0.009919881820678711\n",
      "fold: 3\n",
      "train_r2: 0.2182593308119979 test_r2: 0.12837241047315484 independent_r2: 0.017331897555470377 runtime: 0.007919788360595703\n",
      "fold: 4\n",
      "train_r2: 0.21461938224655164 test_r2: 0.25304538768031815 independent_r2: 0.010162358225582091 runtime: 0.012205362319946289\n",
      "fold: 5\n",
      "train_r2: 0.22083790340914966 test_r2: 0.09829646761496758 independent_r2: 0.006601227825036804 runtime: 0.009419918060302734\n",
      "fold: 6\n",
      "train_r2: 0.22328790195054848 test_r2: 0.0529039393785985 independent_r2: -0.015625057436272627 runtime: 0.007543087005615234\n",
      "fold: 7\n",
      "train_r2: 0.22776062381308881 test_r2: 0.13838410244718458 independent_r2: 0.02795153705480713 runtime: 0.011033058166503906\n",
      "fold: 8\n",
      "train_r2: 0.25131120038887045 test_r2: -0.07261427072925297 independent_r2: -0.013121692297399568 runtime: 0.01627039909362793\n",
      "fold: 9\n",
      "train_r2: 0.22012614157896704 test_r2: 0.12969329401251517 independent_r2: -0.00870490774478272 runtime: 0.01742839813232422\n",
      "Mut KernelRidge ic50\n",
      "Mut\n",
      "fold: 0\n",
      "train_r2: 0.20254791274193773 test_r2: -0.03810399731564651 independent_r2: 0.04348534683644112 runtime: 0.010604143142700195\n",
      "fold: 1\n",
      "train_r2: 0.18327681133235985 test_r2: 0.006679029637074119 independent_r2: 0.05822606318787005 runtime: 0.011258840560913086\n",
      "fold: 2\n",
      "train_r2: 0.18557943996984483 test_r2: -0.010454134349365596 independent_r2: 0.01669911740834784 runtime: 0.010352611541748047\n",
      "fold: 3\n",
      "train_r2: 0.17916114759899604 test_r2: 0.08969403840186863 independent_r2: 0.030797348073936748 runtime: 0.018533706665039062\n",
      "fold: 4\n",
      "train_r2: 0.18785214342812684 test_r2: 0.07553381205956022 independent_r2: 0.05181936492704309 runtime: 0.02411651611328125\n",
      "fold: 5\n",
      "train_r2: 0.17333421005236926 test_r2: 0.17462265377759167 independent_r2: 0.046514515116631516 runtime: 0.031310081481933594\n",
      "fold: 6\n",
      "train_r2: 0.1895864333049252 test_r2: -0.18813448738730387 independent_r2: 0.014174127226505906 runtime: 0.020546674728393555\n",
      "fold: 7\n",
      "train_r2: 0.19154179401325344 test_r2: 0.08798204956600808 independent_r2: 0.04487690213964013 runtime: 0.02285003662109375\n",
      "fold: 8\n",
      "train_r2: 0.18752115601758124 test_r2: 0.07694394569753604 independent_r2: 0.024528335255996514 runtime: 0.021683692932128906\n",
      "fold: 9\n",
      "train_r2: 0.18269336426588734 test_r2: 0.08604989405053642 independent_r2: 0.03143388411929826 runtime: 0.015899181365966797\n",
      "VAF XGBoost auc\n",
      "VAF\n",
      "fold: 0\n",
      "train_r2: 0.6390834745040201 test_r2: -1.0827205132774815 independent_r2: -0.22435126058758237 runtime: 0.17044377326965332\n",
      "fold: 1\n",
      "train_r2: 0.6804105645424348 test_r2: -0.8537512271296597 independent_r2: -0.26751290213688406 runtime: 0.04882979393005371\n",
      "fold: 2\n",
      "train_r2: 0.6590643599785957 test_r2: -0.46544857477656665 independent_r2: -0.34764807115758023 runtime: 0.045090675354003906\n",
      "fold: 3\n",
      "train_r2: 0.6647745667682388 test_r2: -0.6165951623167183 independent_r2: -0.29450892612494783 runtime: 0.03918766975402832\n",
      "fold: 4\n",
      "train_r2: 0.6403168007000399 test_r2: -0.4862851067438847 independent_r2: -0.4219897227975289 runtime: 0.04831051826477051\n",
      "fold: 5\n",
      "train_r2: 0.6477179485174083 test_r2: -0.5952917870716803 independent_r2: -0.2307520082530472 runtime: 0.06536746025085449\n",
      "fold: 6\n",
      "train_r2: 0.64897955312861 test_r2: -0.4746803429684634 independent_r2: -0.3014389078719697 runtime: 0.050499677658081055\n",
      "fold: 7\n",
      "train_r2: 0.6857123677889292 test_r2: -0.12126191899898298 independent_r2: -0.36536581453106565 runtime: 0.03935384750366211\n",
      "fold: 8\n",
      "train_r2: 0.6347706987327837 test_r2: -0.7282186373455364 independent_r2: -0.2469979214791096 runtime: 0.06391358375549316\n",
      "fold: 9\n",
      "train_r2: 0.6722157570846345 test_r2: -0.1319652570279728 independent_r2: -0.29345534002466533 runtime: 0.0683126449584961\n",
      "VAF XGBoost ic50\n",
      "VAF\n",
      "fold: 0\n",
      "train_r2: 0.6663118443139081 test_r2: -0.7590045437193926 independent_r2: -0.37906639988932467 runtime: 0.06173396110534668\n",
      "fold: 1\n",
      "train_r2: 0.7171850780942945 test_r2: -1.203059083633049 independent_r2: -0.4537274480175304 runtime: 0.4638485908508301\n",
      "fold: 2\n",
      "train_r2: 0.68453885427096 test_r2: -0.3162024853953729 independent_r2: -0.4882975026400811 runtime: 0.04872918128967285\n",
      "fold: 3\n",
      "train_r2: 0.6922709218258569 test_r2: -0.34047749698509655 independent_r2: -0.578678320003011 runtime: 0.04497790336608887\n",
      "fold: 4\n",
      "train_r2: 0.6678541611442335 test_r2: -0.5151144185493237 independent_r2: -0.5492964738739812 runtime: 0.0492556095123291\n",
      "fold: 5\n",
      "train_r2: 0.6805338466827691 test_r2: -0.180054884586774 independent_r2: -0.3498380814766824 runtime: 0.04819798469543457\n",
      "fold: 6\n",
      "train_r2: 0.6966329645908353 test_r2: -2.3450131384486554 independent_r2: -0.7025938111597079 runtime: 0.07060980796813965\n",
      "fold: 7\n",
      "train_r2: 0.7077735664810322 test_r2: -0.21225072315376736 independent_r2: -0.5100060329998155 runtime: 0.04831290245056152\n",
      "fold: 8\n",
      "train_r2: 0.6608519677647067 test_r2: -0.7612175942356612 independent_r2: -0.4941356814248916 runtime: 0.04024910926818848\n",
      "fold: 9\n",
      "train_r2: 0.6787024012962619 test_r2: -0.3474804460533687 independent_r2: -0.5675975860672005 runtime: 0.04185009002685547\n",
      "VAF LassoCV auc\n",
      "VAF\n",
      "fold: 0\n",
      "train_r2: 0.21124706225280987 test_r2: -0.1920609785673797 independent_r2: -0.05061072429221647 runtime: 0.049138545989990234\n",
      "fold: 1\n",
      "train_r2: 0.23326169857435541 test_r2: -0.4457109203076155 independent_r2: -0.09043048602684944 runtime: 0.05181169509887695\n",
      "fold: 2\n",
      "train_r2: 0.18436200193377883 test_r2: 0.1514759446121512 independent_r2: -0.04537014823361307 runtime: 0.0565037727355957\n",
      "fold: 3\n",
      "train_r2: 0.20745132442054093 test_r2: 0.047218193063595426 independent_r2: -0.09221747204547937 runtime: 0.049149274826049805\n",
      "fold: 4\n",
      "train_r2: 0.218873309256584 test_r2: -0.244902096696074 independent_r2: -0.07440931928041139 runtime: 0.05182647705078125\n",
      "fold: 5\n",
      "train_r2: 0.19424647838286369 test_r2: 0.13412298937117584 independent_r2: -0.04756678545365434 runtime: 0.05087137222290039\n",
      "fold: 6\n",
      "train_r2: 0.1703927074322703 test_r2: 0.06291665758197729 independent_r2: -0.020675792260550274 runtime: 0.047919511795043945\n",
      "fold: 7\n",
      "train_r2: 0.2130305833090006 test_r2: 0.019178449408526532 independent_r2: -0.05091942927263138 runtime: 0.04529714584350586\n",
      "fold: 8\n",
      "train_r2: 0.19445515500908617 test_r2: 0.14301297402746815 independent_r2: -0.07504780004104816 runtime: 0.04784417152404785\n",
      "fold: 9\n",
      "train_r2: 0.19037496674251586 test_r2: 0.13789651794251034 independent_r2: -0.03599306890556009 runtime: 0.04691147804260254\n",
      "VAF LassoCV ic50\n",
      "VAF\n",
      "fold: 0\n",
      "train_r2: 0.15034497078589715 test_r2: -0.03917092959203772 independent_r2: -0.014945158774341438 runtime: 0.05240988731384277\n",
      "fold: 1\n",
      "train_r2: 0.16635733926998908 test_r2: -0.3686808803404067 independent_r2: 0.009602568853695481 runtime: 0.0471343994140625\n",
      "fold: 2\n",
      "train_r2: 0.12493617588007944 test_r2: 0.0005351962301826108 independent_r2: -0.018448727336864312 runtime: 0.047091007232666016\n",
      "fold: 3\n",
      "train_r2: 0.1252627313369482 test_r2: 0.06863992129926866 independent_r2: 0.010483167837872065 runtime: 0.04860687255859375\n",
      "fold: 4\n",
      "train_r2: 0.14488643190700423 test_r2: -0.16600870677253488 independent_r2: 0.026075846849049067 runtime: 0.04711151123046875\n",
      "fold: 5\n",
      "train_r2: 0.1253485278968285 test_r2: 0.02911504523901376 independent_r2: 0.02625620857878308 runtime: 0.053533077239990234\n",
      "fold: 6\n",
      "train_r2: 0.13051506416110992 test_r2: -0.24957702036731377 independent_r2: 0.0018778989159989878 runtime: 0.046903133392333984\n",
      "fold: 7\n",
      "train_r2: 0.1298120664616884 test_r2: 0.10029055787149976 independent_r2: 0.028157824392066155 runtime: 0.04627728462219238\n",
      "fold: 8\n",
      "train_r2: 0.10820912637013536 test_r2: 0.10914552789978826 independent_r2: -0.014468126714194618 runtime: 0.04599809646606445\n",
      "fold: 9\n",
      "train_r2: 0.10611046976417193 test_r2: 0.10386796007362176 independent_r2: -0.002155501682918226 runtime: 0.045645952224731445\n",
      "VAF KernelRidge auc\n",
      "VAF\n",
      "fold: 0\n",
      "train_r2: 0.05927622643043429 test_r2: -0.0645601036733412 independent_r2: -0.0037695631447507427 runtime: 0.02631068229675293\n",
      "fold: 1\n",
      "train_r2: 0.04978490804214175 test_r2: -0.2167127780254987 independent_r2: 0.0037803442984786217 runtime: 0.03257274627685547\n",
      "fold: 2\n",
      "train_r2: 0.04827470597674277 test_r2: -0.04282551613765029 independent_r2: -0.02361722855112025 runtime: 0.019820690155029297\n",
      "fold: 3\n",
      "train_r2: 0.05420839962357604 test_r2: 0.023741383840992225 independent_r2: -0.007552321132204609 runtime: 0.02851128578186035\n",
      "fold: 4\n",
      "train_r2: 0.05752226197298804 test_r2: -0.011137061917467328 independent_r2: -0.004485567001141888 runtime: 0.025833845138549805\n",
      "fold: 5\n",
      "train_r2: 0.04991840826115346 test_r2: -0.024465215882241154 independent_r2: -0.01726626426779765 runtime: 0.02562737464904785\n",
      "fold: 6\n",
      "train_r2: 0.04382617943372391 test_r2: -0.1174400752953455 independent_r2: -0.02698121043588464 runtime: 0.013658285140991211\n",
      "fold: 7\n",
      "train_r2: 0.054879947208755775 test_r2: 0.015909466363436264 independent_r2: -0.004601913744801456 runtime: 0.009252548217773438\n",
      "fold: 8\n",
      "train_r2: 0.059297353964498756 test_r2: -0.01707621866936959 independent_r2: -0.009709729374149934 runtime: 0.02425837516784668\n",
      "fold: 9\n",
      "train_r2: 0.05284805619763955 test_r2: 0.01439639154313066 independent_r2: -0.01720339948481886 runtime: 0.01866316795349121\n",
      "VAF KernelRidge ic50\n",
      "VAF\n",
      "fold: 0\n",
      "train_r2: 0.05049385750771629 test_r2: -0.0037680094221297367 independent_r2: -0.0008252744541066281 runtime: 0.021581172943115234\n",
      "fold: 1\n",
      "train_r2: 0.044910798972946475 test_r2: -0.1740426968327622 independent_r2: 0.012886987282194062 runtime: 0.00944828987121582\n",
      "fold: 2\n",
      "train_r2: 0.04500524421481533 test_r2: -0.07437958754390728 independent_r2: -0.00990495242892453 runtime: 0.015116214752197266\n",
      "fold: 3\n",
      "train_r2: 0.04517469203729019 test_r2: 0.02954503529878716 independent_r2: -0.0002524740155271932 runtime: 0.008131980895996094\n",
      "fold: 4\n",
      "train_r2: 0.05449251846375691 test_r2: -0.06771796546493136 independent_r2: 0.010583997316914062 runtime: 0.010077476501464844\n",
      "fold: 5\n",
      "train_r2: 0.042067528189318804 test_r2: 0.012772019629719411 independent_r2: -0.00190821311555478 runtime: 0.012231826782226562\n",
      "fold: 6\n",
      "train_r2: 0.04467679954238313 test_r2: -0.33271265017272267 independent_r2: -0.018656323229428562 runtime: 0.009650468826293945\n",
      "fold: 7\n",
      "train_r2: 0.04364877655329591 test_r2: -0.007823382477307073 independent_r2: 0.008405188666214647 runtime: 0.016440391540527344\n",
      "fold: 8\n",
      "train_r2: 0.04289469607166985 test_r2: 0.03799150592515055 independent_r2: -0.0024137126413372734 runtime: 0.01597118377685547\n",
      "fold: 9\n",
      "train_r2: 0.043050382663277764 test_r2: 0.030098199040525575 independent_r2: 0.0004275291886902677 runtime: 0.01636505126953125\n",
      "Mut+VAF XGBoost auc\n",
      "Mut+VAF\n",
      "fold: 0\n",
      "train_r2: 0.7962702616056476 test_r2: -0.6449611331508176 independent_r2: -0.40462741937698143 runtime: 0.1802823543548584\n",
      "fold: 1\n",
      "train_r2: 0.8011553048868443 test_r2: -0.26213958936575144 independent_r2: -0.25783563885867977 runtime: 0.05629611015319824\n",
      "fold: 2\n",
      "train_r2: 0.8036723917840387 test_r2: -0.3190862010591713 independent_r2: -0.39086686515908164 runtime: 0.0583956241607666\n",
      "fold: 3\n",
      "train_r2: 0.8215535140523486 test_r2: -0.09091831416034646 independent_r2: -0.3670684484002551 runtime: 0.0548100471496582\n",
      "fold: 4\n",
      "train_r2: 0.8205382081672681 test_r2: -0.5472795176618384 independent_r2: -0.4018921588837818 runtime: 0.07126259803771973\n",
      "fold: 5\n",
      "train_r2: 0.8243747043728877 test_r2: -0.45601729132653746 independent_r2: -0.27129025651181915 runtime: 0.06302857398986816\n",
      "fold: 6\n",
      "train_r2: 0.8090318087685111 test_r2: -0.19416572732972415 independent_r2: -0.359434518498152 runtime: 0.04907965660095215\n",
      "fold: 7\n",
      "train_r2: 0.8348688781750563 test_r2: -0.17193836144755315 independent_r2: -0.4135185691648331 runtime: 0.05037856101989746\n",
      "fold: 8\n",
      "train_r2: 0.7924212898342919 test_r2: -0.8093347924655965 independent_r2: -0.2901711054497129 runtime: 0.0811467170715332\n",
      "fold: 9\n",
      "train_r2: 0.8434402971149393 test_r2: 0.09151995255753287 independent_r2: -0.3112050679340914 runtime: 0.0662546157836914\n",
      "Mut+VAF XGBoost ic50\n",
      "Mut+VAF\n",
      "fold: 0\n",
      "train_r2: 0.8325139231827066 test_r2: -0.4776504691708081 independent_r2: -0.4662847240558954 runtime: 0.06956243515014648\n",
      "fold: 1\n",
      "train_r2: 0.8330481995777509 test_r2: -0.733161415894878 independent_r2: -0.3416149721999022 runtime: 0.046994686126708984\n",
      "fold: 2\n",
      "train_r2: 0.8352078008255978 test_r2: -0.29433472086506307 independent_r2: -0.4873476225281421 runtime: 0.08980727195739746\n",
      "fold: 3\n",
      "train_r2: 0.8421866560360926 test_r2: 0.1210605726805084 independent_r2: -0.6873293189148237 runtime: 0.05482602119445801\n",
      "fold: 4\n",
      "train_r2: 0.8521606816571252 test_r2: -0.34205176205020593 independent_r2: -0.5229437890387154 runtime: 0.07061052322387695\n",
      "fold: 5\n",
      "train_r2: 0.8521330572495885 test_r2: 0.010528660212177687 independent_r2: -0.46815664534925894 runtime: 0.06287884712219238\n",
      "fold: 6\n",
      "train_r2: 0.8489921248090909 test_r2: -0.945705261811667 independent_r2: -0.5106802926188341 runtime: 0.07597684860229492\n",
      "fold: 7\n",
      "train_r2: 0.8774727037767616 test_r2: -0.5177387740426429 independent_r2: -0.5895198123428633 runtime: 0.05044221878051758\n",
      "fold: 8\n",
      "train_r2: 0.8296427345587715 test_r2: -0.36359997316368076 independent_r2: -0.4764769557355111 runtime: 0.0738973617553711\n",
      "fold: 9\n",
      "train_r2: 0.8479076311437926 test_r2: -0.35969796059051906 independent_r2: -0.5168249489576942 runtime: 0.0741426944732666\n",
      "Mut+VAF LassoCV auc\n",
      "Mut+VAF\n",
      "fold: 0\n",
      "train_r2: 0.2800797269624823 test_r2: -0.20771747896330384 independent_r2: -0.00213787202414073 runtime: 0.06075763702392578\n",
      "fold: 1\n",
      "train_r2: 0.2970670793373611 test_r2: 0.0003736556602894625 independent_r2: 0.01979935738367178 runtime: 0.05418825149536133\n",
      "fold: 2\n",
      "train_r2: 0.2720566343460761 test_r2: 0.21685018318723615 independent_r2: -0.03971862984444474 runtime: 0.04925823211669922\n",
      "fold: 3\n",
      "train_r2: 0.3213859830206476 test_r2: 0.22531472877035463 independent_r2: -0.041021388591898855 runtime: 0.051572322845458984\n",
      "fold: 4\n",
      "train_r2: 0.321092047843088 test_r2: 0.187104777637497 independent_r2: -0.036082099354728614 runtime: 0.05070829391479492\n",
      "fold: 5\n",
      "train_r2: 0.3009252425623492 test_r2: 0.07628343461836173 independent_r2: -0.025205133942797886 runtime: 0.05721139907836914\n",
      "fold: 6\n",
      "train_r2: 0.32815614050487063 test_r2: -0.07428213521615845 independent_r2: -0.05835279654123915 runtime: 0.0530853271484375\n",
      "fold: 7\n",
      "train_r2: 0.31712465572714255 test_r2: 0.13966152811637766 independent_r2: 0.00448171939711417 runtime: 0.06089639663696289\n",
      "fold: 8\n",
      "train_r2: 0.2985155390049772 test_r2: -0.2692259185857002 independent_r2: -0.05021231533243675 runtime: 0.057685136795043945\n",
      "fold: 9\n",
      "train_r2: 0.22249287283202746 test_r2: 0.16013073075190865 independent_r2: -0.033614343049533035 runtime: 0.059446096420288086\n",
      "Mut+VAF LassoCV ic50\n",
      "Mut+VAF\n",
      "fold: 0\n",
      "train_r2: 0.23525084975240407 test_r2: -0.15246824590044383 independent_r2: 0.04316543075308654 runtime: 0.05315136909484863\n",
      "fold: 1\n",
      "train_r2: 0.20466586124305253 test_r2: -0.06804299284121762 independent_r2: 0.06791738186877172 runtime: 0.05814790725708008\n",
      "fold: 2\n",
      "train_r2: 0.19370038147252333 test_r2: 0.07004174069309377 independent_r2: -0.0027027782557729196 runtime: 0.050927162170410156\n",
      "fold: 3\n",
      "train_r2: 0.19307519769305748 test_r2: 0.15524960222195383 independent_r2: 0.040637800357385845 runtime: 0.0485532283782959\n",
      "fold: 4\n",
      "train_r2: 0.22806608451118549 test_r2: -0.03135490151885478 independent_r2: 0.043119499169859354 runtime: 0.05453228950500488\n",
      "fold: 5\n",
      "train_r2: 0.235650766539251 test_r2: 0.12336562577726162 independent_r2: 0.00489079110376156 runtime: 0.05436539649963379\n",
      "fold: 6\n",
      "train_r2: 0.2126815304522669 test_r2: -0.33319239466623696 independent_r2: 0.0010172694529431148 runtime: 0.055629730224609375\n",
      "fold: 7\n",
      "train_r2: 0.25595504918245937 test_r2: -0.010966373174071675 independent_r2: 0.0035214987481618776 runtime: 0.05325818061828613\n",
      "fold: 8\n",
      "train_r2: 0.22397016518245383 test_r2: 0.06549206680069575 independent_r2: -0.002342510699721023 runtime: 0.06857776641845703\n",
      "fold: 9\n",
      "train_r2: 0.15981881806511833 test_r2: 0.06832836718947333 independent_r2: -0.004532055996475437 runtime: 0.06574010848999023\n",
      "Mut+VAF KernelRidge auc\n",
      "Mut+VAF\n",
      "fold: 0\n",
      "train_r2: 0.20255271642599992 test_r2: -0.021891077526907665 independent_r2: 0.007064837128647694 runtime: 0.01607823371887207\n",
      "fold: 1\n",
      "train_r2: 0.18599419818845897 test_r2: -0.06372369884955909 independent_r2: 0.02498993231595492 runtime: 0.019118547439575195\n",
      "fold: 2\n",
      "train_r2: 0.1828205822536012 test_r2: 0.08894935890240707 independent_r2: -0.020938616753374273 runtime: 0.01455998420715332\n",
      "fold: 3\n",
      "train_r2: 0.1827569261729589 test_r2: 0.17963993707958392 independent_r2: 0.009363590437220126 runtime: 0.02099752426147461\n",
      "fold: 4\n",
      "train_r2: 0.1826611689751344 test_r2: 0.19807584533004008 independent_r2: 0.0052596090807934814 runtime: 0.024092912673950195\n",
      "fold: 5\n",
      "train_r2: 0.18841725971765966 test_r2: 0.07888029958876475 independent_r2: -0.006968822461236979 runtime: 0.029367923736572266\n",
      "fold: 6\n",
      "train_r2: 0.18972441760516745 test_r2: -0.03751258921734646 independent_r2: -0.016544616608864526 runtime: 0.01692342758178711\n",
      "fold: 7\n",
      "train_r2: 0.1879928550945471 test_r2: 0.14258623798998515 independent_r2: 0.01696985011117169 runtime: 0.013761758804321289\n",
      "fold: 8\n",
      "train_r2: 0.22374655224562723 test_r2: -0.17802195638977225 independent_r2: -0.018335141503388774 runtime: 0.015388965606689453\n",
      "fold: 9\n",
      "train_r2: 0.19076230796980043 test_r2: 0.0997608359529506 independent_r2: -0.012826789827056873 runtime: 0.02765345573425293\n",
      "Mut+VAF KernelRidge ic50\n",
      "Mut+VAF\n",
      "fold: 0\n",
      "train_r2: 0.16925476279729923 test_r2: -0.013975959039714825 independent_r2: 0.031174150803091383 runtime: 0.029789209365844727\n",
      "fold: 1\n",
      "train_r2: 0.152850705617529 test_r2: -0.07777886377997367 independent_r2: 0.048465747133069836 runtime: 0.02964949607849121\n",
      "fold: 2\n",
      "train_r2: 0.1543191462005704 test_r2: -0.016043969584037532 independent_r2: 0.003620623219195984 runtime: 0.03644108772277832\n",
      "fold: 3\n",
      "train_r2: 0.15170011371728642 test_r2: 0.08904567114706696 independent_r2: 0.02392601839372943 runtime: 0.026293039321899414\n",
      "fold: 4\n",
      "train_r2: 0.15629406106408494 test_r2: 0.029007092368802834 independent_r2: 0.04307963256078784 runtime: 0.02752208709716797\n",
      "fold: 5\n",
      "train_r2: 0.14585780478859212 test_r2: 0.12175767461703224 independent_r2: 0.02984106131366948 runtime: 0.02823948860168457\n",
      "fold: 6\n",
      "train_r2: 0.15781056988633368 test_r2: -0.2608238961831364 independent_r2: 0.0031580025334947415 runtime: 0.03738141059875488\n",
      "fold: 7\n",
      "train_r2: 0.15893742518080856 test_r2: 0.042441720010420014 independent_r2: 0.039280674080480193 runtime: 0.026309967041015625\n",
      "fold: 8\n",
      "train_r2: 0.1576102451304845 test_r2: 0.055114714303124535 independent_r2: 0.015232502092353384 runtime: 0.025717496871948242\n",
      "fold: 9\n",
      "train_r2: 0.15404324962475802 test_r2: 0.08066012655338373 independent_r2: 0.021652199488965573 runtime: 0.02084517478942871\n",
      "Expr XGBoost auc\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9999999999884072 test_r2: 0.49655110247177636 independent_r2: 0.43955254470934635 runtime: 8.918163537979126\n",
      "fold: 1\n",
      "train_r2: 0.9999999999879373 test_r2: 0.5498601625962524 independent_r2: 0.4039493955554828 runtime: 9.488836765289307\n",
      "fold: 2\n",
      "train_r2: 0.999999999988684 test_r2: 0.6120837760234464 independent_r2: 0.32639535286753674 runtime: 12.435246229171753\n",
      "fold: 3\n",
      "train_r2: 0.9999999999879893 test_r2: 0.2870591746977126 independent_r2: 0.4205980518292384 runtime: 10.994618654251099\n",
      "fold: 4\n",
      "train_r2: 0.9999999999888327 test_r2: 0.5874245638812969 independent_r2: 0.3357515371433276 runtime: 11.20065951347351\n",
      "fold: 5\n",
      "train_r2: 0.9999999999875714 test_r2: 0.33805494584049545 independent_r2: 0.1775868483751536 runtime: 11.34067177772522\n",
      "fold: 6\n",
      "train_r2: 0.9999999999883898 test_r2: 0.4680623142161654 independent_r2: 0.215789515352512 runtime: 11.993335247039795\n",
      "fold: 7\n",
      "train_r2: 0.9999999999884337 test_r2: 0.2877430007918448 independent_r2: 0.22540216527596824 runtime: 12.206376075744629\n",
      "fold: 8\n",
      "train_r2: 0.9999999999864349 test_r2: 0.567997362418416 independent_r2: 0.36001185351873966 runtime: 12.039322137832642\n",
      "fold: 9\n",
      "train_r2: 0.9999999999867859 test_r2: 0.413454383113785 independent_r2: 0.30055559819559985 runtime: 12.583354234695435\n",
      "Expr XGBoost ic50\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.9999999968634886 test_r2: -0.07801227219428752 independent_r2: 0.054382574357096036 runtime: 10.627995491027832\n",
      "fold: 1\n",
      "train_r2: 0.999999997263479 test_r2: 0.37339821383029603 independent_r2: 0.24564355693273132 runtime: 10.933385610580444\n",
      "fold: 2\n",
      "train_r2: 0.9999999966342116 test_r2: 0.4161575668299765 independent_r2: 0.24579914277678439 runtime: 11.04453706741333\n",
      "fold: 3\n",
      "train_r2: 0.9999999969959615 test_r2: 0.27541105450095416 independent_r2: -0.10339602882173948 runtime: 13.630505561828613\n",
      "fold: 4\n",
      "train_r2: 0.9999999971141053 test_r2: 0.3689815450595978 independent_r2: 0.016707486078291867 runtime: 12.23389720916748\n",
      "fold: 5\n",
      "train_r2: 0.9999999970612462 test_r2: 0.14660557853271639 independent_r2: 0.10320607463436637 runtime: 11.366177082061768\n",
      "fold: 6\n",
      "train_r2: 0.9999999969765425 test_r2: 0.47588785672651146 independent_r2: 0.08727380770053783 runtime: 11.041863918304443\n",
      "fold: 7\n",
      "train_r2: 0.999999996867414 test_r2: 0.07864156184661408 independent_r2: 0.06788310873707293 runtime: 11.64127779006958\n",
      "fold: 8\n",
      "train_r2: 0.9999999972300367 test_r2: 0.43282351046337775 independent_r2: 0.0837294729414726 runtime: 11.843852758407593\n",
      "fold: 9\n",
      "train_r2: 0.9999999969800205 test_r2: -0.1684080571253035 independent_r2: -0.13965517713983333 runtime: 12.328428745269775\n",
      "Expr LassoCV auc\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.7780514703426211 test_r2: 0.2995760186316735 independent_r2: 0.38486045989322193 runtime: 31.623106002807617\n",
      "fold: 1\n",
      "train_r2: 0.7590630424364729 test_r2: 0.690724702601051 independent_r2: 0.39514388629629926 runtime: 29.905318021774292\n",
      "fold: 2\n",
      "train_r2: 0.7147727179915623 test_r2: 0.6782649442801554 independent_r2: 0.4692420023275068 runtime: 38.97229743003845\n",
      "fold: 3\n",
      "train_r2: 0.7035853984174051 test_r2: 0.7263930213236331 independent_r2: 0.47397399275133156 runtime: 31.910142421722412\n",
      "fold: 4\n",
      "train_r2: 0.7059390209715033 test_r2: 0.6328011785828507 independent_r2: 0.4451991888965343 runtime: 29.352097511291504\n",
      "fold: 5\n",
      "train_r2: 0.6795066162873147 test_r2: 0.6759342260467769 independent_r2: 0.4702715120604918 runtime: 31.56818461418152\n",
      "fold: 6\n",
      "train_r2: 0.7216225232309588 test_r2: 0.5818862166852035 independent_r2: 0.40524651119956345 runtime: 30.772249937057495\n",
      "fold: 7\n",
      "train_r2: 0.7095806463040839 test_r2: 0.49347775184131515 independent_r2: 0.5137621574798072 runtime: 33.05256938934326\n",
      "fold: 8\n",
      "train_r2: 0.8581755055264272 test_r2: 0.414607509460738 independent_r2: 0.39264553220114107 runtime: 31.17721438407898\n",
      "fold: 9\n",
      "train_r2: 0.7252410890198355 test_r2: 0.3537434940261607 independent_r2: 0.4934050468939486 runtime: 31.77306866645813\n",
      "Expr LassoCV ic50\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.46076845545544265 test_r2: 0.05775042927557672 independent_r2: 0.26966004429892154 runtime: 34.22100639343262\n",
      "fold: 1\n",
      "train_r2: 0.4793261769541083 test_r2: 0.417266832791915 independent_r2: 0.23747677126350353 runtime: 33.684744119644165\n",
      "fold: 2\n",
      "train_r2: 0.7185566627207378 test_r2: 0.28892834719504956 independent_r2: 0.2607563981176806 runtime: 34.799113750457764\n",
      "fold: 3\n",
      "train_r2: 0.49950714263804386 test_r2: 0.1854863764162734 independent_r2: 0.2951309012149611 runtime: 36.63799810409546\n",
      "fold: 4\n",
      "train_r2: 0.4676293707496446 test_r2: 0.4927826360015768 independent_r2: 0.29427889410100483 runtime: 32.89087963104248\n",
      "fold: 5\n",
      "train_r2: 0.41764091569633444 test_r2: 0.3653743055384979 independent_r2: 0.31722865976393355 runtime: 40.0147750377655\n",
      "fold: 6\n",
      "train_r2: 0.6175592483593124 test_r2: 0.15885109843848177 independent_r2: 0.22262307651803204 runtime: 34.856696367263794\n",
      "fold: 7\n",
      "train_r2: 0.6136083946137868 test_r2: 0.06626185016879094 independent_r2: 0.3109702087628824 runtime: 39.71686768531799\n",
      "fold: 8\n",
      "train_r2: 0.43331553975817605 test_r2: 0.38988622959874275 independent_r2: 0.2756473903673753 runtime: 43.97718930244446\n",
      "fold: 9\n",
      "train_r2: 0.5616632224814342 test_r2: 0.15013575778380894 independent_r2: 0.26757588788187137 runtime: 38.66815686225891\n",
      "Expr KernelRidge auc\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.7901163093613489 test_r2: 0.11207493805827584 independent_r2: -0.9150885902442312 runtime: 0.8499538898468018\n",
      "fold: 1\n",
      "train_r2: 0.7827558131651694 test_r2: 0.3597630623768887 independent_r2: -0.930853397446453 runtime: 0.8875339031219482\n",
      "fold: 2\n",
      "train_r2: 0.7827082286654974 test_r2: 0.5651604201290867 independent_r2: -0.8898713245781245 runtime: 0.8962388038635254\n",
      "fold: 3\n",
      "train_r2: 0.791322565369666 test_r2: 0.3298009748193581 independent_r2: -0.8937644307922228 runtime: 0.9202852249145508\n",
      "fold: 4\n",
      "train_r2: 0.7890222391574108 test_r2: 0.29341336771944937 independent_r2: -0.9265165537846045 runtime: 0.8652029037475586\n",
      "fold: 5\n",
      "train_r2: 0.7872043313974317 test_r2: 0.51811352479323 independent_r2: -0.9238185248749635 runtime: 0.872687578201294\n",
      "fold: 6\n",
      "train_r2: 0.7893559478887203 test_r2: 0.30128065463808384 independent_r2: -0.9296134419411968 runtime: 0.8912696838378906\n",
      "fold: 7\n",
      "train_r2: 0.7878785496357474 test_r2: 0.23641364296613898 independent_r2: -0.9116518979884276 runtime: 0.8521521091461182\n",
      "fold: 8\n",
      "train_r2: 0.7853875138898478 test_r2: 0.5369967481146853 independent_r2: -0.9261998626115193 runtime: 0.9579370021820068\n",
      "fold: 9\n",
      "train_r2: 0.8017683536050677 test_r2: 0.033733935285083394 independent_r2: -0.949597442524492 runtime: 1.011756420135498\n",
      "Expr KernelRidge ic50\n",
      "Expr\n",
      "fold: 0\n",
      "train_r2: 0.7780384690614933 test_r2: 0.22129031954742218 independent_r2: -0.032917665473145785 runtime: 0.9377050399780273\n",
      "fold: 1\n",
      "train_r2: 0.7711587636569643 test_r2: 0.3210592154048504 independent_r2: -0.04716282401989136 runtime: 0.9187190532684326\n",
      "fold: 2\n",
      "train_r2: 0.7781646986840117 test_r2: 0.28681483104014505 independent_r2: -0.028634238242599697 runtime: 0.9070312976837158\n",
      "fold: 3\n",
      "train_r2: 0.7841661833898141 test_r2: 0.06361628175175948 independent_r2: -0.02904329348660495 runtime: 0.8721017837524414\n",
      "fold: 4\n",
      "train_r2: 0.7680734851108726 test_r2: 0.44917309832936547 independent_r2: -0.04108250270743086 runtime: 0.8809471130371094\n",
      "fold: 5\n",
      "train_r2: 0.7694977970866856 test_r2: 0.42629800556606234 independent_r2: -0.05719664519402157 runtime: 0.8807117938995361\n",
      "fold: 6\n",
      "train_r2: 0.7721952845457984 test_r2: 0.31399475581149405 independent_r2: -0.02915959189352768 runtime: 0.8915205001831055\n",
      "fold: 7\n",
      "train_r2: 0.7829893858173759 test_r2: 0.11081464994126156 independent_r2: -0.034967900455322054 runtime: 0.8948655128479004\n",
      "fold: 8\n",
      "train_r2: 0.7668824549107975 test_r2: 0.5167718008740514 independent_r2: -0.06457971543457286 runtime: 0.8896727561950684\n",
      "fold: 9\n",
      "train_r2: 0.7823630539655082 test_r2: 0.050729612710801675 independent_r2: -0.0555650959174836 runtime: 0.9187467098236084\n",
      "RelativeExpr XGBoost auc\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9999999998725184 test_r2: 0.20436557735310368 independent_r2: 0.5488915633756679 runtime: 0.6392421722412109\n",
      "fold: 1\n",
      "train_r2: 0.9999999999220466 test_r2: 0.5660985218577618 independent_r2: 0.5252268727425895 runtime: 0.2306079864501953\n",
      "fold: 2\n",
      "train_r2: 0.9999999998606681 test_r2: 0.6989360662286521 independent_r2: 0.5454295239985758 runtime: 0.32071375846862793\n",
      "fold: 3\n",
      "train_r2: 0.9999999998700593 test_r2: 0.349542745353173 independent_r2: 0.5212936191147501 runtime: 0.2991924285888672\n",
      "fold: 4\n",
      "train_r2: 0.9999999998751496 test_r2: 0.6352675544037274 independent_r2: 0.5248117398533894 runtime: 0.26326870918273926\n",
      "fold: 5\n",
      "train_r2: 0.9999999999056293 test_r2: 0.48594077121073553 independent_r2: 0.6133405731888244 runtime: 0.25514864921569824\n",
      "fold: 6\n",
      "train_r2: 0.9999999998958147 test_r2: 0.7899238121564647 independent_r2: 0.5058652751101013 runtime: 0.2681906223297119\n",
      "fold: 7\n",
      "train_r2: 0.9999999999030317 test_r2: 0.6998332433030514 independent_r2: 0.5383892177714242 runtime: 0.24463891983032227\n",
      "fold: 8\n",
      "train_r2: 0.9999999998773226 test_r2: 0.6870329298931106 independent_r2: 0.5754523771120038 runtime: 0.5436012744903564\n",
      "fold: 9\n",
      "train_r2: 0.9999999998660617 test_r2: 0.3343564551223649 independent_r2: 0.5399088623479738 runtime: 0.23797202110290527\n",
      "RelativeExpr XGBoost ic50\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9999999734680165 test_r2: 0.13900617573468987 independent_r2: 0.31344873052810285 runtime: 0.18172836303710938\n",
      "fold: 1\n",
      "train_r2: 0.9999999758200686 test_r2: 0.23702221634129172 independent_r2: 0.335649224486827 runtime: 0.30181193351745605\n",
      "fold: 2\n",
      "train_r2: 0.9999999758374813 test_r2: 0.30131553411943246 independent_r2: 0.3666352526462018 runtime: 0.20044159889221191\n",
      "fold: 3\n",
      "train_r2: 0.9999999729644614 test_r2: 0.15927050522923736 independent_r2: 0.36080754941985205 runtime: 0.1845700740814209\n",
      "fold: 4\n",
      "train_r2: 0.9999999718153744 test_r2: 0.47228367288631656 independent_r2: 0.34719626914152957 runtime: 0.22132587432861328\n",
      "fold: 5\n",
      "train_r2: 0.9999999763383333 test_r2: 0.1878659828776318 independent_r2: 0.35959372078060003 runtime: 0.1963033676147461\n",
      "fold: 6\n",
      "train_r2: 0.9999999765045373 test_r2: -0.04646511073038395 independent_r2: 0.22956355085249447 runtime: 0.19107341766357422\n",
      "fold: 7\n",
      "train_r2: 0.9999999807547411 test_r2: 0.4317205494019105 independent_r2: 0.34111565471451677 runtime: 0.2246990203857422\n",
      "fold: 8\n",
      "train_r2: 0.9999999795606064 test_r2: 0.5834983750962185 independent_r2: 0.301444819839759 runtime: 0.7039978504180908\n",
      "fold: 9\n",
      "train_r2: 0.9999999663706479 test_r2: 0.12329916619560577 independent_r2: 0.3231627247227765 runtime: 0.1707751750946045\n",
      "RelativeExpr LassoCV auc\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.8321825481366398 test_r2: 0.3219709932951793 independent_r2: 0.5830147115400481 runtime: 0.47684454917907715\n",
      "fold: 1\n",
      "train_r2: 0.8374911523375707 test_r2: 0.6490013543620792 independent_r2: 0.5889771910124515 runtime: 0.6073296070098877\n",
      "fold: 2\n",
      "train_r2: 0.8084380742792736 test_r2: 0.7569021612922686 independent_r2: 0.6019629993555122 runtime: 0.8037607669830322\n",
      "fold: 3\n",
      "train_r2: 0.808802445237866 test_r2: 0.6948620550141756 independent_r2: 0.591458731698548 runtime: 0.7752323150634766\n",
      "fold: 4\n",
      "train_r2: 0.7671223806525204 test_r2: 0.6780252201955945 independent_r2: 0.5981043848754928 runtime: 0.6272194385528564\n",
      "fold: 5\n",
      "train_r2: 0.8234162289775298 test_r2: 0.5959582158122676 independent_r2: 0.6128664349142064 runtime: 0.6106109619140625\n",
      "fold: 6\n",
      "train_r2: 0.792303724749612 test_r2: 0.7556577424462472 independent_r2: 0.5817782509816014 runtime: 1.0683605670928955\n",
      "fold: 7\n",
      "train_r2: 0.762891547923122 test_r2: 0.5624425262844219 independent_r2: 0.5974851220883752 runtime: 0.7470271587371826\n",
      "fold: 8\n",
      "train_r2: 0.852780820327383 test_r2: 0.676018914710127 independent_r2: 0.5794253932364689 runtime: 0.5140953063964844\n",
      "fold: 9\n",
      "train_r2: 0.8521752117182284 test_r2: 0.49430024444066023 independent_r2: 0.6008729793749659 runtime: 0.6020829677581787\n",
      "RelativeExpr LassoCV ic50\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.5900918007396826 test_r2: 0.07168401831322391 independent_r2: 0.43500886487858836 runtime: 0.9868073463439941\n",
      "fold: 1\n",
      "train_r2: 0.6293992956055376 test_r2: 0.5241975115022235 independent_r2: 0.39264611094220914 runtime: 0.7105441093444824\n",
      "fold: 2\n",
      "train_r2: 0.6251447033373143 test_r2: 0.4867951228745523 independent_r2: 0.3949494125085079 runtime: 0.966482400894165\n",
      "fold: 3\n",
      "train_r2: 0.5829434380551957 test_r2: 0.19984877661497547 independent_r2: 0.41229785773585736 runtime: 0.8783769607543945\n",
      "fold: 4\n",
      "train_r2: 0.5487915648502656 test_r2: 0.5581134743339402 independent_r2: 0.42314729630803094 runtime: 1.0189182758331299\n",
      "fold: 5\n",
      "train_r2: 0.6062781905840313 test_r2: 0.33163492625241187 independent_r2: 0.4067491711216784 runtime: 0.85723876953125\n",
      "fold: 6\n",
      "train_r2: 0.5553525486036259 test_r2: 0.3122573329985199 independent_r2: 0.39979385597772576 runtime: 0.7745094299316406\n",
      "fold: 7\n",
      "train_r2: 0.579727551534138 test_r2: 0.22633642127687048 independent_r2: 0.4165795065598977 runtime: 0.630246639251709\n",
      "fold: 8\n",
      "train_r2: 0.5661596563814777 test_r2: 0.4828648188275487 independent_r2: 0.43383664316463144 runtime: 0.7295784950256348\n",
      "fold: 9\n",
      "train_r2: 0.6621586262553933 test_r2: 0.25665669302583694 independent_r2: 0.38902843539682175 runtime: 0.5434129238128662\n",
      "RelativeExpr KernelRidge auc\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.7561811403395124 test_r2: 0.40330485544800887 independent_r2: 0.5986523886644577 runtime: 0.06708383560180664\n",
      "fold: 1\n",
      "train_r2: 0.7268846206920645 test_r2: 0.7748830422063872 independent_r2: 0.5997833090302866 runtime: 0.03103470802307129\n",
      "fold: 2\n",
      "train_r2: 0.7360358619230725 test_r2: 0.7334523421077608 independent_r2: 0.5973631674435536 runtime: 0.03521108627319336\n",
      "fold: 3\n",
      "train_r2: 0.7378436994897108 test_r2: 0.652527994996916 independent_r2: 0.5899271036745073 runtime: 0.030254364013671875\n",
      "fold: 4\n",
      "train_r2: 0.7361553266388262 test_r2: 0.7198793740217873 independent_r2: 0.5963848758832002 runtime: 0.04058718681335449\n",
      "fold: 5\n",
      "train_r2: 0.7469214783619442 test_r2: 0.6101122205581848 independent_r2: 0.604225488887302 runtime: 0.035195112228393555\n",
      "fold: 6\n",
      "train_r2: 0.7311789910224682 test_r2: 0.7672464111629561 independent_r2: 0.5898608216351758 runtime: 0.038604736328125\n",
      "fold: 7\n",
      "train_r2: 0.7433980535090079 test_r2: 0.5988898575134356 independent_r2: 0.5962790214480216 runtime: 0.07893157005310059\n",
      "fold: 8\n",
      "train_r2: 0.7380360001993393 test_r2: 0.6832297281341715 independent_r2: 0.6033247448108543 runtime: 0.08027005195617676\n",
      "fold: 9\n",
      "train_r2: 0.7653174066710848 test_r2: 0.46581716196612677 independent_r2: 0.6023034337812313 runtime: 0.07287859916687012\n",
      "RelativeExpr KernelRidge ic50\n",
      "RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.614963627132814 test_r2: 0.20326318387644537 independent_r2: 0.42852226299901663 runtime: 0.061286211013793945\n",
      "fold: 1\n",
      "train_r2: 0.5780839786620413 test_r2: 0.583205048435996 independent_r2: 0.4299701842275072 runtime: 0.07168269157409668\n",
      "fold: 2\n",
      "train_r2: 0.5885693525435047 test_r2: 0.4790084361112845 independent_r2: 0.43747130764186715 runtime: 0.06559133529663086\n",
      "fold: 3\n",
      "train_r2: 0.60452769952249 test_r2: 0.21363036974969585 independent_r2: 0.42117085756599104 runtime: 0.03646111488342285\n",
      "fold: 4\n",
      "train_r2: 0.5790663275851307 test_r2: 0.6238224863175343 independent_r2: 0.44008823085276605 runtime: 0.02998971939086914\n",
      "fold: 5\n",
      "train_r2: 0.6021154273875422 test_r2: 0.3691875721530158 independent_r2: 0.43729797119589897 runtime: 0.028664588928222656\n",
      "fold: 6\n",
      "train_r2: 0.589607104601769 test_r2: 0.39471589703491716 independent_r2: 0.4279415914374397 runtime: 0.028189420700073242\n",
      "fold: 7\n",
      "train_r2: 0.6057203537793174 test_r2: 0.345302689547376 independent_r2: 0.4428734033334729 runtime: 0.04012346267700195\n",
      "fold: 8\n",
      "train_r2: 0.5891256680307314 test_r2: 0.5178003562145143 independent_r2: 0.44099170379173247 runtime: 0.05525517463684082\n",
      "fold: 9\n",
      "train_r2: 0.5999864931971155 test_r2: 0.34370034943584216 independent_r2: 0.4500960209308218 runtime: 0.05703234672546387\n",
      "Mut+RelativeExpr XGBoost auc\n",
      "Mut+RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9999999999293464 test_r2: 0.3618758994228498 independent_r2: 0.6298494654031284 runtime: 0.3671114444732666\n",
      "fold: 1\n",
      "train_r2: 0.999999999902074 test_r2: 0.5133334145816116 independent_r2: 0.6004939605341708 runtime: 0.27289581298828125\n",
      "fold: 2\n",
      "train_r2: 0.9999999998999236 test_r2: 0.6955513156961237 independent_r2: 0.5366247110236861 runtime: 0.2359769344329834\n",
      "fold: 3\n",
      "train_r2: 0.9999999998887157 test_r2: 0.5793015757244162 independent_r2: 0.5820658260634535 runtime: 0.22999835014343262\n",
      "fold: 4\n",
      "train_r2: 0.999999999863991 test_r2: 0.6735322961016124 independent_r2: 0.49029715238494076 runtime: 0.2111046314239502\n",
      "fold: 5\n",
      "train_r2: 0.9999999999067418 test_r2: 0.2083829695766638 independent_r2: 0.5881911313605835 runtime: 0.3461766242980957\n",
      "fold: 6\n",
      "train_r2: 0.9999999999191977 test_r2: 0.6662175188019135 independent_r2: 0.5477809355831943 runtime: 0.2533872127532959\n",
      "fold: 7\n",
      "train_r2: 0.9999999998493166 test_r2: 0.6487154023292476 independent_r2: 0.5320524025068065 runtime: 0.31864476203918457\n",
      "fold: 8\n",
      "train_r2: 0.9999999999003109 test_r2: 0.6559582597660472 independent_r2: 0.5963205473868354 runtime: 0.3305211067199707\n",
      "fold: 9\n",
      "train_r2: 0.9999999999090026 test_r2: 0.4005631263377497 independent_r2: 0.5993743338862362 runtime: 0.28586578369140625\n",
      "Mut+RelativeExpr XGBoost ic50\n",
      "Mut+RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9999999788763072 test_r2: -0.1140053157801304 independent_r2: 0.33588680613179756 runtime: 0.3356661796569824\n",
      "fold: 1\n",
      "train_r2: 0.9999999777718636 test_r2: -0.06659629549709067 independent_r2: 0.3734611969831584 runtime: 0.2167515754699707\n",
      "fold: 2\n",
      "train_r2: 0.9999999733701974 test_r2: 0.16079954605982483 independent_r2: 0.31746848475924927 runtime: 0.35748958587646484\n",
      "fold: 3\n",
      "train_r2: 0.9999999767346839 test_r2: 0.23793174969439612 independent_r2: 0.3221150370311694 runtime: 0.2727510929107666\n",
      "fold: 4\n",
      "train_r2: 0.9999999782750916 test_r2: 0.35925927946879366 independent_r2: 0.3884292225030266 runtime: 0.2375636100769043\n",
      "fold: 5\n",
      "train_r2: 0.9999999814953429 test_r2: 0.026159178294355212 independent_r2: 0.3259935483017954 runtime: 0.3192882537841797\n",
      "fold: 6\n",
      "train_r2: 0.9999999686829826 test_r2: 0.24361540668741355 independent_r2: 0.22988573347154706 runtime: 0.2623124122619629\n",
      "fold: 7\n",
      "train_r2: 0.9999999762436103 test_r2: 0.273495731154643 independent_r2: 0.40542742003310583 runtime: 0.21959495544433594\n",
      "fold: 8\n",
      "train_r2: 0.9999999776431342 test_r2: 0.18083056009825982 independent_r2: 0.27573954180997984 runtime: 0.2438509464263916\n",
      "fold: 9\n",
      "train_r2: 0.9999999665299911 test_r2: 0.4209308282195443 independent_r2: 0.2623861114361189 runtime: 0.2718350887298584\n",
      "Mut+RelativeExpr LassoCV auc\n",
      "Mut+RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.8250140372337191 test_r2: 0.28953268398384224 independent_r2: 0.6004637822194737 runtime: 0.38529133796691895\n",
      "fold: 1\n",
      "train_r2: 0.8569417302809731 test_r2: 0.6107636990635736 independent_r2: 0.5806750991870545 runtime: 1.5968599319458008\n",
      "fold: 2\n",
      "train_r2: 0.8025352153311804 test_r2: 0.7770584330342387 independent_r2: 0.6032525755596878 runtime: 1.1279921531677246\n",
      "fold: 3\n",
      "train_r2: 0.811871775811543 test_r2: 0.6496690265347945 independent_r2: 0.5944827969671324 runtime: 0.6432890892028809\n",
      "fold: 4\n",
      "train_r2: 0.7473533400579555 test_r2: 0.7376016020203149 independent_r2: 0.6037590762875472 runtime: 0.5961639881134033\n",
      "fold: 5\n",
      "train_r2: 0.834201586912138 test_r2: 0.4638625666280499 independent_r2: 0.6144664076892319 runtime: 0.5533773899078369\n",
      "fold: 6\n",
      "train_r2: 0.8007967257957967 test_r2: 0.7275509952851633 independent_r2: 0.583545614917189 runtime: 0.5610008239746094\n",
      "fold: 7\n",
      "train_r2: 0.7751264517737513 test_r2: 0.5867446695495444 independent_r2: 0.5945049006204296 runtime: 0.5481791496276855\n",
      "fold: 8\n",
      "train_r2: 0.7199104738284713 test_r2: 0.6332787422088612 independent_r2: 0.6000302051008835 runtime: 0.514976978302002\n",
      "fold: 9\n",
      "train_r2: 0.8202197606297992 test_r2: 0.47366339503279 independent_r2: 0.6194603666131113 runtime: 0.5759117603302002\n",
      "Mut+RelativeExpr LassoCV ic50\n",
      "Mut+RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.6137042398814958 test_r2: 0.09981747957406228 independent_r2: 0.4235437665981857 runtime: 0.6618902683258057\n",
      "fold: 1\n",
      "train_r2: 0.6797509204622718 test_r2: 0.4671358896420733 independent_r2: 0.33142372587551727 runtime: 0.6141335964202881\n",
      "fold: 2\n",
      "train_r2: 0.6837382781278853 test_r2: 0.5226862375703236 independent_r2: 0.34736565006412745 runtime: 0.6424117088317871\n",
      "fold: 3\n",
      "train_r2: 0.5743520218643379 test_r2: 0.19071164733525015 independent_r2: 0.39761130891311336 runtime: 0.8610758781433105\n",
      "fold: 4\n",
      "train_r2: 0.6467057496439402 test_r2: 0.5990645112259326 independent_r2: 0.36319913677019233 runtime: 0.6194555759429932\n",
      "fold: 5\n",
      "train_r2: 0.7031263381141355 test_r2: 0.15948927092202547 independent_r2: 0.36367770469570493 runtime: 0.6633038520812988\n",
      "fold: 6\n",
      "train_r2: 0.6166039757022151 test_r2: 0.2994032564453746 independent_r2: 0.3668530916981454 runtime: 0.6839776039123535\n",
      "fold: 7\n",
      "train_r2: 0.6539800789518916 test_r2: 0.2472050309838516 independent_r2: 0.3850793872898578 runtime: 0.6447758674621582\n",
      "fold: 8\n",
      "train_r2: 0.7359727169161029 test_r2: 0.44565598043672927 independent_r2: 0.3366186932010057 runtime: 1.3960332870483398\n",
      "fold: 9\n",
      "train_r2: 0.6745951887199038 test_r2: 0.3656820664063902 independent_r2: 0.3581992499192771 runtime: 0.8992383480072021\n",
      "Mut+RelativeExpr KernelRidge auc\n",
      "Mut+RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.7581253090630571 test_r2: 0.3908125574587459 independent_r2: 0.6075703639439893 runtime: 0.041837453842163086\n",
      "fold: 1\n",
      "train_r2: 0.7294639803596301 test_r2: 0.7571747312846153 independent_r2: 0.6059118854488764 runtime: 0.03686261177062988\n",
      "fold: 2\n",
      "train_r2: 0.7353346095118902 test_r2: 0.7519686691622363 independent_r2: 0.6070910866701962 runtime: 0.040007829666137695\n",
      "fold: 3\n",
      "train_r2: 0.7403544483316628 test_r2: 0.6746678424835777 independent_r2: 0.6017220582002553 runtime: 0.05747032165527344\n",
      "fold: 4\n",
      "train_r2: 0.7333696529371815 test_r2: 0.7574708489580049 independent_r2: 0.6052088688940787 runtime: 0.17330193519592285\n",
      "fold: 5\n",
      "train_r2: 0.7527346410452478 test_r2: 0.5463440511189055 independent_r2: 0.612741021732222 runtime: 0.07384610176086426\n",
      "fold: 6\n",
      "train_r2: 0.7330341255393374 test_r2: 0.760636249955111 independent_r2: 0.5975043475615356 runtime: 0.13288116455078125\n",
      "fold: 7\n",
      "train_r2: 0.7435398668897641 test_r2: 0.5988140956703003 independent_r2: 0.6054290624454073 runtime: 0.07810139656066895\n",
      "fold: 8\n",
      "train_r2: 0.7418964562729942 test_r2: 0.6608327831446688 independent_r2: 0.6099440401675877 runtime: 0.11681509017944336\n",
      "fold: 9\n",
      "train_r2: 0.7624295517157789 test_r2: 0.512866263950531 independent_r2: 0.6106441357167058 runtime: 0.17435765266418457\n",
      "Mut+RelativeExpr KernelRidge ic50\n",
      "Mut+RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.6262514902026043 test_r2: 0.18944236742363052 independent_r2: 0.4358426907103917 runtime: 0.15433955192565918\n",
      "fold: 1\n",
      "train_r2: 0.5932939044650495 test_r2: 0.5373904243480788 independent_r2: 0.4251787720840211 runtime: 0.18441462516784668\n",
      "fold: 2\n",
      "train_r2: 0.594671997022447 test_r2: 0.5164822435438596 independent_r2: 0.4351583876120426 runtime: 0.1504828929901123\n",
      "fold: 3\n",
      "train_r2: 0.6157412606067423 test_r2: 0.2540812436782732 independent_r2: 0.4287309042603342 runtime: 0.12176036834716797\n",
      "fold: 4\n",
      "train_r2: 0.5815993282715679 test_r2: 0.6902994529302243 independent_r2: 0.4394988439920985 runtime: 0.12938332557678223\n",
      "fold: 5\n",
      "train_r2: 0.6164848917155077 test_r2: 0.3163890074026845 independent_r2: 0.44124447969394276 runtime: 0.12204241752624512\n",
      "fold: 6\n",
      "train_r2: 0.5956797927433539 test_r2: 0.39588693052783563 independent_r2: 0.42606104118877663 runtime: 0.07439494132995605\n",
      "fold: 7\n",
      "train_r2: 0.6120486087575931 test_r2: 0.36005114776423897 independent_r2: 0.44384106646927934 runtime: 0.06154775619506836\n",
      "fold: 8\n",
      "train_r2: 0.5993980130464555 test_r2: 0.5264474161501267 independent_r2: 0.44094536027336106 runtime: 0.05276799201965332\n",
      "fold: 9\n",
      "train_r2: 0.6021232720110565 test_r2: 0.43178971890864826 independent_r2: 0.4494250630115745 runtime: 0.12514305114746094\n",
      "Mut+VAF+RelativeExpr XGBoost auc\n",
      "Mut+VAF+RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.999999999898036 test_r2: 0.2945394229196394 independent_r2: 0.6278715691796899 runtime: 0.48078060150146484\n",
      "fold: 1\n",
      "train_r2: 0.999999999888553 test_r2: 0.6268144229580406 independent_r2: 0.5649578368387459 runtime: 0.43697524070739746\n",
      "fold: 2\n",
      "train_r2: 0.99999999989845 test_r2: 0.686249814751038 independent_r2: 0.543284502949585 runtime: 0.4266963005065918\n",
      "fold: 3\n",
      "train_r2: 0.9999999999032788 test_r2: 0.5214800180152778 independent_r2: 0.5774392312658971 runtime: 0.24798989295959473\n",
      "fold: 4\n",
      "train_r2: 0.9999999999243282 test_r2: 0.7892136202735689 independent_r2: 0.5470948219021957 runtime: 0.7183008193969727\n",
      "fold: 5\n",
      "train_r2: 0.9999999998965624 test_r2: 0.1644649261514589 independent_r2: 0.5995905798273942 runtime: 0.29640960693359375\n",
      "fold: 6\n",
      "train_r2: 0.9999999998967412 test_r2: 0.7700500982074533 independent_r2: 0.5521241646348016 runtime: 0.5153188705444336\n",
      "fold: 7\n",
      "train_r2: 0.999999999902024 test_r2: 0.4889208350716686 independent_r2: 0.5450259625371034 runtime: 0.38530945777893066\n",
      "fold: 8\n",
      "train_r2: 0.9999999999066145 test_r2: 0.6935073014545423 independent_r2: 0.5958642818686966 runtime: 0.2918248176574707\n",
      "fold: 9\n",
      "train_r2: 0.9999999999050326 test_r2: 0.42648142328688965 independent_r2: 0.5853698067312874 runtime: 0.40429067611694336\n",
      "Mut+VAF+RelativeExpr XGBoost ic50\n",
      "Mut+VAF+RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.9999999774132916 test_r2: 0.19984337391702478 independent_r2: 0.3474711123531731 runtime: 0.30013227462768555\n",
      "fold: 1\n",
      "train_r2: 0.9999999804380918 test_r2: 0.08688445121803279 independent_r2: 0.29796998339998426 runtime: 0.20592427253723145\n",
      "fold: 2\n",
      "train_r2: 0.9999999743304121 test_r2: 0.17808936453711965 independent_r2: 0.3314108561580762 runtime: 0.23483514785766602\n",
      "fold: 3\n",
      "train_r2: 0.9999999837604482 test_r2: 0.33213048638386733 independent_r2: 0.3564276844099632 runtime: 0.3678736686706543\n",
      "fold: 4\n",
      "train_r2: 0.9999999772061103 test_r2: 0.46779444805642434 independent_r2: 0.4074355709165425 runtime: 0.6464078426361084\n",
      "fold: 5\n",
      "train_r2: 0.9999999795451554 test_r2: -0.17022443407790178 independent_r2: 0.4010485031277885 runtime: 0.531928539276123\n",
      "fold: 6\n",
      "train_r2: 0.9999999753708358 test_r2: 0.366905078377859 independent_r2: 0.281908573860703 runtime: 0.646838903427124\n",
      "fold: 7\n",
      "train_r2: 0.9999999712844528 test_r2: -0.009599038935060644 independent_r2: 0.3636926875866209 runtime: 0.3762331008911133\n",
      "fold: 8\n",
      "train_r2: 0.9999999811600119 test_r2: 0.5031810582968488 independent_r2: 0.30549220061049986 runtime: 0.2303638458251953\n",
      "fold: 9\n",
      "train_r2: 0.9999999724709048 test_r2: 0.18534721790648023 independent_r2: 0.3216526535353462 runtime: 0.2855210304260254\n",
      "Mut+VAF+RelativeExpr LassoCV auc\n",
      "Mut+VAF+RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.795130701939945 test_r2: 0.2811838367596894 independent_r2: 0.6030947542113173 runtime: 0.8776302337646484\n",
      "fold: 1\n",
      "train_r2: 0.8177537771005231 test_r2: 0.602546564440462 independent_r2: 0.594218673540958 runtime: 1.1318235397338867\n",
      "fold: 2\n",
      "train_r2: 0.8036117002648044 test_r2: 0.7497207909829458 independent_r2: 0.5964591566851752 runtime: 2.769650459289551\n",
      "fold: 3\n",
      "train_r2: 0.8034381742938066 test_r2: 0.6209861591953498 independent_r2: 0.5934722811763969 runtime: 1.1309056282043457\n",
      "fold: 4\n",
      "train_r2: 0.726846931953208 test_r2: 0.6924342168913756 independent_r2: 0.6030292746944932 runtime: 0.7600581645965576\n",
      "fold: 5\n",
      "train_r2: 0.8755528129256669 test_r2: 0.3558092286598111 independent_r2: 0.5821802426578715 runtime: 0.9861533641815186\n",
      "fold: 6\n",
      "train_r2: 0.8010463719210742 test_r2: 0.683511571161201 independent_r2: 0.5774211135919383 runtime: 0.6724207401275635\n",
      "fold: 7\n",
      "train_r2: 0.784174697568512 test_r2: 0.44089553447326923 independent_r2: 0.5948216235230874 runtime: 0.5080232620239258\n",
      "fold: 8\n",
      "train_r2: 0.7576258057321927 test_r2: 0.6442262849098235 independent_r2: 0.6100261193999506 runtime: 0.7122151851654053\n",
      "fold: 9\n",
      "train_r2: 0.8020682125086337 test_r2: 0.4636471849348409 independent_r2: 0.6139848923471255 runtime: 0.6959960460662842\n",
      "Mut+VAF+RelativeExpr LassoCV ic50\n",
      "Mut+VAF+RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.613166574948764 test_r2: 0.10434567394131078 independent_r2: 0.41527153989052445 runtime: 0.6529879570007324\n",
      "fold: 1\n",
      "train_r2: 0.630758737117139 test_r2: 0.48403163988438513 independent_r2: 0.34625999835978427 runtime: 1.3220500946044922\n",
      "fold: 2\n",
      "train_r2: 0.6765950277345079 test_r2: 0.4785325326160782 independent_r2: 0.34898641208909764 runtime: 1.4466896057128906\n",
      "fold: 3\n",
      "train_r2: 0.5801382667094671 test_r2: 0.23579207477478092 independent_r2: 0.39278466602983086 runtime: 1.2096898555755615\n",
      "fold: 4\n",
      "train_r2: 0.6529682264950154 test_r2: 0.5806071997408327 independent_r2: 0.3472117682256268 runtime: 1.438999891281128\n",
      "fold: 5\n",
      "train_r2: 0.6935650889995091 test_r2: 0.1593871758167975 independent_r2: 0.36450558810243183 runtime: 1.1543657779693604\n",
      "fold: 6\n",
      "train_r2: 0.6322891255618155 test_r2: 0.37805332834036687 independent_r2: 0.3614410336688255 runtime: 1.382110834121704\n",
      "fold: 7\n",
      "train_r2: 0.6381807895112931 test_r2: 0.22921617323646204 independent_r2: 0.38642826755886517 runtime: 2.901146173477173\n",
      "fold: 8\n",
      "train_r2: 0.730085914031728 test_r2: 0.4470600479086597 independent_r2: 0.3427871780205969 runtime: 0.8250610828399658\n",
      "fold: 9\n",
      "train_r2: 0.666131238192034 test_r2: 0.37522034026520945 independent_r2: 0.35819037895262607 runtime: 0.8596248626708984\n",
      "Mut+VAF+RelativeExpr KernelRidge auc\n",
      "Mut+VAF+RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.7437289078524526 test_r2: 0.3879169546511124 independent_r2: 0.6006874497687297 runtime: 0.09884166717529297\n",
      "fold: 1\n",
      "train_r2: 0.7142783749327709 test_r2: 0.7290855270189385 independent_r2: 0.598821980460963 runtime: 0.11914777755737305\n",
      "fold: 2\n",
      "train_r2: 0.7208733311599045 test_r2: 0.7283885285414418 independent_r2: 0.5993794509453023 runtime: 0.06223273277282715\n",
      "fold: 3\n",
      "train_r2: 0.7271356332650922 test_r2: 0.6307849618866903 independent_r2: 0.5964997036564175 runtime: 0.06306982040405273\n",
      "fold: 4\n",
      "train_r2: 0.717029247813902 test_r2: 0.7483058687701765 independent_r2: 0.5966635479188935 runtime: 0.1269822120666504\n",
      "fold: 5\n",
      "train_r2: 0.7362402309927449 test_r2: 0.5563079671754816 independent_r2: 0.6056215886923485 runtime: 0.10357427597045898\n",
      "fold: 6\n",
      "train_r2: 0.716864589679191 test_r2: 0.7526880080904027 independent_r2: 0.5883337122379506 runtime: 0.09343910217285156\n",
      "fold: 7\n",
      "train_r2: 0.7340749202457348 test_r2: 0.5153067648058169 independent_r2: 0.6015286982393573 runtime: 0.05785560607910156\n",
      "fold: 8\n",
      "train_r2: 0.7256252229109644 test_r2: 0.6586069074162324 independent_r2: 0.6018153739807375 runtime: 0.04458904266357422\n",
      "fold: 9\n",
      "train_r2: 0.74715488737725 test_r2: 0.5092227967396672 independent_r2: 0.6023323520459252 runtime: 0.043776750564575195\n",
      "Mut+VAF+RelativeExpr KernelRidge ic50\n",
      "Mut+VAF+RelativeExpr\n",
      "fold: 0\n",
      "train_r2: 0.614810306395106 test_r2: 0.18970244832943028 independent_r2: 0.4326744450325012 runtime: 0.11094093322753906\n",
      "fold: 1\n",
      "train_r2: 0.5797100643250888 test_r2: 0.5298276225309728 independent_r2: 0.4216977473004946 runtime: 0.11104989051818848\n",
      "fold: 2\n",
      "train_r2: 0.5881278054421957 test_r2: 0.4628847694695426 independent_r2: 0.4305084020424055 runtime: 0.10859084129333496\n",
      "fold: 3\n",
      "train_r2: 0.6033572814373842 test_r2: 0.2697934921216829 independent_r2: 0.4346058139475226 runtime: 0.09060907363891602\n",
      "fold: 4\n",
      "train_r2: 0.5676365354253488 test_r2: 0.7261823258738191 independent_r2: 0.43226105983286645 runtime: 0.08945727348327637\n",
      "fold: 5\n",
      "train_r2: 0.6062392752368679 test_r2: 0.30122231383804554 independent_r2: 0.4385683250960981 runtime: 0.10933494567871094\n",
      "fold: 6\n",
      "train_r2: 0.5810387448339782 test_r2: 0.4872954901884091 independent_r2: 0.423533734222317 runtime: 0.13381505012512207\n",
      "fold: 7\n",
      "train_r2: 0.6020085006008582 test_r2: 0.32700283382838424 independent_r2: 0.43963723382406983 runtime: 0.12194514274597168\n",
      "fold: 8\n",
      "train_r2: 0.5871488675207885 test_r2: 0.5268149808561273 independent_r2: 0.4376896544036408 runtime: 0.06525087356567383\n",
      "fold: 9\n",
      "train_r2: 0.5901640001276784 test_r2: 0.43223961758182616 independent_r2: 0.44655412396463623 runtime: 0.06011247634887695\n"
     ]
    }
   ],
   "source": [
    "# TODO: we can use the thresholds solely for balancing our k-fold cross-validations.\n",
    "# or, we don't use the thresholds at all, and just do 10-fold cross-validation.\n",
    "\n",
    "import time\n",
    "regression_results = {}\n",
    "results_columns = []\n",
    "for dataset, method, target in itertools.product(feature_sets, methods, targets):\n",
    "    print(dataset, method, target)\n",
    "    result_all = run_regression_dataset(dataset, dic_dataset_train, dic_dataset_test, method=method, target=target)\n",
    "    regression_results[(dataset, method, target)] = result_all"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "regression_results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Run every drug"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 290,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_drug_regression_results = {}\n",
    "all_data = pd.read_csv('all_drug_regression_results.csv', index_col=None)\n",
    "for i, row in all_data.iterrows():\n",
    "    k = (row.drug, row.feature_set, row.method, row.target)\n",
    "    if k not in all_drug_regression_results:\n",
    "        all_drug_regression_results[k] = {}\n",
    "    row_cols = row.keys()[4:]\n",
    "    row_dict = {}\n",
    "    for c in row_cols:\n",
    "        row_dict[c] = row[c]\n",
    "    all_drug_regression_results[k][row.iter] = row_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 295,
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_sets = ['Mut', 'VAF', 'Mut+VAF', 'Expr', 'RelativeExpr', 'Mut+RelativeExpr', 'Mut+VAF+RelativeExpr']\n",
    "methods = ['XGBoost', 'LassoCV', 'KernelRidge']\n",
    "targets = ['auc', 'ic50']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "for drug in all_drugs.to_list():\n",
    "    print(drug)\n",
    "    w12_f123 = BeatAML_w12_data.loc[:,featureset_1 + featureset_2 + featureset_3]\n",
    "    w12_drug = BeatAML_w12_data.loc[:,[drug, drug+'_auc']]\n",
    "    df_sele_w12 = pd.concat([ w12_f123,new_feature_df_t,w12_drug],axis = 1)\n",
    "\n",
    "    w34_f123 = BeatAML_w34_data.loc[:,featureset_1 + featureset_2 + featureset_3]\n",
    "    w34_drug = BeatAML_w34_data.loc[:,[drug, drug+'_auc']]\n",
    "    df_sele_w34 = pd.concat([ w34_f123,new_feature_df_test,w34_drug],axis = 1)\n",
    "    \n",
    "    dic_dataset_test_drug = {\n",
    "    \"Mut\": df_sele_w34[set1 + [drug, drug+'_auc']].dropna(),\n",
    "    \"VAF\": df_sele_w34[set2 + [drug, drug+'_auc']].dropna(),\n",
    "    \"Mut+VAF\": df_sele_w34[set1 + set2 + [drug, drug+'_auc']].dropna(),\n",
    "    \"Expr\": df_sele_w34[set3 + [drug, drug+'_auc']].dropna(),\n",
    "    \"RelativeExpr\": df_sele_w34[set4 + [drug, drug+'_auc']].dropna(),\n",
    "    \"Mut+RelativeExpr\": df_sele_w34[set1+ set4 + [drug, drug+'_auc']].dropna(),\n",
    "    \"Mut+VAF+RelativeExpr\": df_sele_w34[set1 + set2 + set4 + [drug, drug+'_auc']].dropna()\n",
    "    }\n",
    "\n",
    "    dic_dataset_train_drug = {\n",
    "    \"Mut\": df_sele_w12[set1 + [drug, drug+'_auc']].dropna(),\n",
    "    \"VAF\": df_sele_w12[set2 + [drug, drug+'_auc']].dropna(),\n",
    "    \"Mut+VAF\": df_sele_w12[set1 + set2 + [drug, drug+'_auc']].dropna(),\n",
    "    \"Expr\": pd.concat([df_sele_w12[set3], df_sele_w12[[drug, drug+'_auc']]], axis = 1).dropna(),\n",
    "    \"RelativeExpr\": df_sele_w12[set4 + [drug, drug+'_auc']].dropna(),\n",
    "    \"Mut+RelativeExpr\": df_sele_w12[set1+ set4 + [drug, drug+'_auc']].dropna(),\n",
    "    \"Mut+VAF+RelativeExpr\": df_sele_w12[set1 + set2 + set4 + [drug, drug+'_auc']].dropna()\n",
    "    }\n",
    "    for dataset, method, target in itertools.product(feature_sets, methods, targets):\n",
    "        print(drug, dataset, method, target)\n",
    "        if (drug, dataset, method, target) in all_drug_regression_results and (method != 'KernelRidge' or dataset != 'Expr'):\n",
    "            old_results = all_drug_regression_results[(drug, dataset, method, target)]\n",
    "            print('Already ran')\n",
    "            continue\n",
    "            for k, v in old_results.items():\n",
    "                print(k, 'train_r2:', old_results[k]['train_r2'], 'test_r2:', old_results[k]['test_r2'],\n",
    "                  'independent_r2:', old_results[k]['independent_r2'], 'runtime:', old_results[k]['runtime'])\n",
    "            continue\n",
    "        result_all = run_regression_dataset(dataset, dic_dataset_train_drug, dic_dataset_test_drug, method=method, target=target,\n",
    "                                           drug=drug)\n",
    "        all_drug_regression_results[(drug, dataset, method, target)] = result_all"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 305,
   "metadata": {},
   "outputs": [],
   "source": [
    "results_columns = [x for x in all_drug_regression_results[('Venetoclax', 'Mut', 'XGBoost', 'auc')][0].keys() if x!='features' and x!='iter']\n",
    "# create a data table for venetoclax only\n",
    "# columns: drug, method, target, features, run\n",
    "all_drug_data_table_columns = ['drug', 'feature_set', 'method', 'target', 'iter'] + results_columns\n",
    "all_drug_data_table = []\n",
    "for key, results in all_drug_regression_results.items():\n",
    "    for i, result in results.items():\n",
    "        row = list(key) + [i]\n",
    "        for c in results_columns:\n",
    "            row.append(result[c])\n",
    "        all_drug_data_table.append(row)\n",
    "all_drug_data_table = pd.DataFrame(all_drug_data_table)\n",
    "all_drug_data_table.columns = all_drug_data_table_columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 307,
   "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>drug</th>\n",
       "      <th>feature_set</th>\n",
       "      <th>method</th>\n",
       "      <th>target</th>\n",
       "      <th>iter</th>\n",
       "      <th>train_mse</th>\n",
       "      <th>train_mape</th>\n",
       "      <th>train_r2</th>\n",
       "      <th>test_mse</th>\n",
       "      <th>test_mape</th>\n",
       "      <th>test_r2</th>\n",
       "      <th>independent_mse</th>\n",
       "      <th>independent_mape</th>\n",
       "      <th>independent_r2</th>\n",
       "      <th>runtime</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Idelalisib</td>\n",
       "      <td>Mut</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>auc</td>\n",
       "      <td>0</td>\n",
       "      <td>959.992369</td>\n",
       "      <td>0.129452</td>\n",
       "      <td>0.581095</td>\n",
       "      <td>1966.460221</td>\n",
       "      <td>0.203349</td>\n",
       "      <td>-0.318271</td>\n",
       "      <td>3170.229831</td>\n",
       "      <td>0.279143</td>\n",
       "      <td>-0.460721</td>\n",
       "      <td>0.104214</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Idelalisib</td>\n",
       "      <td>Mut</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>auc</td>\n",
       "      <td>1</td>\n",
       "      <td>960.930242</td>\n",
       "      <td>0.124880</td>\n",
       "      <td>0.579634</td>\n",
       "      <td>2627.605214</td>\n",
       "      <td>0.260969</td>\n",
       "      <td>-0.721189</td>\n",
       "      <td>3217.808723</td>\n",
       "      <td>0.285678</td>\n",
       "      <td>-0.482643</td>\n",
       "      <td>0.078411</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Idelalisib</td>\n",
       "      <td>Mut</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>auc</td>\n",
       "      <td>2</td>\n",
       "      <td>920.367785</td>\n",
       "      <td>0.122378</td>\n",
       "      <td>0.555459</td>\n",
       "      <td>3869.717785</td>\n",
       "      <td>0.982554</td>\n",
       "      <td>-0.188009</td>\n",
       "      <td>3134.253765</td>\n",
       "      <td>0.284423</td>\n",
       "      <td>-0.444144</td>\n",
       "      <td>0.182187</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Idelalisib</td>\n",
       "      <td>Mut</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>auc</td>\n",
       "      <td>3</td>\n",
       "      <td>1002.184000</td>\n",
       "      <td>0.130755</td>\n",
       "      <td>0.562346</td>\n",
       "      <td>2523.918806</td>\n",
       "      <td>0.297630</td>\n",
       "      <td>-0.660336</td>\n",
       "      <td>2912.155572</td>\n",
       "      <td>0.269998</td>\n",
       "      <td>-0.341810</td>\n",
       "      <td>0.046575</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Idelalisib</td>\n",
       "      <td>Mut</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>auc</td>\n",
       "      <td>4</td>\n",
       "      <td>991.416719</td>\n",
       "      <td>0.130646</td>\n",
       "      <td>0.561705</td>\n",
       "      <td>2991.424357</td>\n",
       "      <td>0.221167</td>\n",
       "      <td>-0.715383</td>\n",
       "      <td>3092.059362</td>\n",
       "      <td>0.279164</td>\n",
       "      <td>-0.424703</td>\n",
       "      <td>0.051976</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",
       "      <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>26455</th>\n",
       "      <td>NF-kB Activation Inhibitor</td>\n",
       "      <td>Mut+VAF+RelativeExpr</td>\n",
       "      <td>KernelRidge</td>\n",
       "      <td>ic50</td>\n",
       "      <td>5</td>\n",
       "      <td>11.180819</td>\n",
       "      <td>23.613386</td>\n",
       "      <td>0.354530</td>\n",
       "      <td>10.623985</td>\n",
       "      <td>0.623880</td>\n",
       "      <td>-0.147664</td>\n",
       "      <td>18.172167</td>\n",
       "      <td>62.690129</td>\n",
       "      <td>0.050743</td>\n",
       "      <td>0.106537</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26456</th>\n",
       "      <td>NF-kB Activation Inhibitor</td>\n",
       "      <td>Mut+VAF+RelativeExpr</td>\n",
       "      <td>KernelRidge</td>\n",
       "      <td>ic50</td>\n",
       "      <td>6</td>\n",
       "      <td>10.599231</td>\n",
       "      <td>21.989947</td>\n",
       "      <td>0.360651</td>\n",
       "      <td>17.795076</td>\n",
       "      <td>20.829713</td>\n",
       "      <td>-0.027042</td>\n",
       "      <td>18.845979</td>\n",
       "      <td>65.030530</td>\n",
       "      <td>0.015545</td>\n",
       "      <td>0.115442</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26457</th>\n",
       "      <td>NF-kB Activation Inhibitor</td>\n",
       "      <td>Mut+VAF+RelativeExpr</td>\n",
       "      <td>KernelRidge</td>\n",
       "      <td>ic50</td>\n",
       "      <td>7</td>\n",
       "      <td>10.894789</td>\n",
       "      <td>18.777758</td>\n",
       "      <td>0.352143</td>\n",
       "      <td>13.968149</td>\n",
       "      <td>61.220142</td>\n",
       "      <td>0.111303</td>\n",
       "      <td>19.206829</td>\n",
       "      <td>65.067884</td>\n",
       "      <td>-0.003305</td>\n",
       "      <td>0.086179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26458</th>\n",
       "      <td>NF-kB Activation Inhibitor</td>\n",
       "      <td>Mut+VAF+RelativeExpr</td>\n",
       "      <td>KernelRidge</td>\n",
       "      <td>ic50</td>\n",
       "      <td>8</td>\n",
       "      <td>10.932295</td>\n",
       "      <td>19.653808</td>\n",
       "      <td>0.332883</td>\n",
       "      <td>14.994918</td>\n",
       "      <td>52.970202</td>\n",
       "      <td>0.228154</td>\n",
       "      <td>18.846284</td>\n",
       "      <td>65.242372</td>\n",
       "      <td>0.015529</td>\n",
       "      <td>0.077884</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26459</th>\n",
       "      <td>NF-kB Activation Inhibitor</td>\n",
       "      <td>Mut+VAF+RelativeExpr</td>\n",
       "      <td>KernelRidge</td>\n",
       "      <td>ic50</td>\n",
       "      <td>9</td>\n",
       "      <td>10.859036</td>\n",
       "      <td>22.766470</td>\n",
       "      <td>0.348994</td>\n",
       "      <td>14.164582</td>\n",
       "      <td>14.584882</td>\n",
       "      <td>0.165418</td>\n",
       "      <td>18.904211</td>\n",
       "      <td>64.549626</td>\n",
       "      <td>0.012503</td>\n",
       "      <td>0.090321</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>26460 rows × 15 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                             drug           feature_set       method target  \\\n",
       "0                      Idelalisib                   Mut      XGBoost    auc   \n",
       "1                      Idelalisib                   Mut      XGBoost    auc   \n",
       "2                      Idelalisib                   Mut      XGBoost    auc   \n",
       "3                      Idelalisib                   Mut      XGBoost    auc   \n",
       "4                      Idelalisib                   Mut      XGBoost    auc   \n",
       "...                           ...                   ...          ...    ...   \n",
       "26455  NF-kB Activation Inhibitor  Mut+VAF+RelativeExpr  KernelRidge   ic50   \n",
       "26456  NF-kB Activation Inhibitor  Mut+VAF+RelativeExpr  KernelRidge   ic50   \n",
       "26457  NF-kB Activation Inhibitor  Mut+VAF+RelativeExpr  KernelRidge   ic50   \n",
       "26458  NF-kB Activation Inhibitor  Mut+VAF+RelativeExpr  KernelRidge   ic50   \n",
       "26459  NF-kB Activation Inhibitor  Mut+VAF+RelativeExpr  KernelRidge   ic50   \n",
       "\n",
       "       iter    train_mse  train_mape  train_r2     test_mse  test_mape  \\\n",
       "0         0   959.992369    0.129452  0.581095  1966.460221   0.203349   \n",
       "1         1   960.930242    0.124880  0.579634  2627.605214   0.260969   \n",
       "2         2   920.367785    0.122378  0.555459  3869.717785   0.982554   \n",
       "3         3  1002.184000    0.130755  0.562346  2523.918806   0.297630   \n",
       "4         4   991.416719    0.130646  0.561705  2991.424357   0.221167   \n",
       "...     ...          ...         ...       ...          ...        ...   \n",
       "26455     5    11.180819   23.613386  0.354530    10.623985   0.623880   \n",
       "26456     6    10.599231   21.989947  0.360651    17.795076  20.829713   \n",
       "26457     7    10.894789   18.777758  0.352143    13.968149  61.220142   \n",
       "26458     8    10.932295   19.653808  0.332883    14.994918  52.970202   \n",
       "26459     9    10.859036   22.766470  0.348994    14.164582  14.584882   \n",
       "\n",
       "        test_r2  independent_mse  independent_mape  independent_r2   runtime  \n",
       "0     -0.318271      3170.229831          0.279143       -0.460721  0.104214  \n",
       "1     -0.721189      3217.808723          0.285678       -0.482643  0.078411  \n",
       "2     -0.188009      3134.253765          0.284423       -0.444144  0.182187  \n",
       "3     -0.660336      2912.155572          0.269998       -0.341810  0.046575  \n",
       "4     -0.715383      3092.059362          0.279164       -0.424703  0.051976  \n",
       "...         ...              ...               ...             ...       ...  \n",
       "26455 -0.147664        18.172167         62.690129        0.050743  0.106537  \n",
       "26456 -0.027042        18.845979         65.030530        0.015545  0.115442  \n",
       "26457  0.111303        19.206829         65.067884       -0.003305  0.086179  \n",
       "26458  0.228154        18.846284         65.242372        0.015529  0.077884  \n",
       "26459  0.165418        18.904211         64.549626        0.012503  0.090321  \n",
       "\n",
       "[26460 rows x 15 columns]"
      ]
     },
     "execution_count": 307,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_drug_data_table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 308,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_drug_data_table.to_csv('all_drug_regression_results.csv', index=None)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Plotting/analyzing results - Venetoclax only"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [],
   "source": [
    "# TODO: plot results for venetoclax only?\n",
    "results_columns = [x for x in regression_results[('Mut', 'XGBoost', 'auc')][0].keys() if x!='features']\n",
    "# create a data table for venetoclax only\n",
    "# columns: drug, method, target, features, run\n",
    "data_table_columns = ['feature_set', 'method', 'target', 'iter'] + results_columns\n",
    "data_table = []\n",
    "for key, results in regression_results.items():\n",
    "    for i, result in results.items():\n",
    "        row = list(key) + [i]\n",
    "        for c in results_columns:\n",
    "            row.append(result[c])\n",
    "        data_table.append(row)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_table = pd.DataFrame(data_table)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_table.columns = data_table_columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "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>feature_set</th>\n",
       "      <th>method</th>\n",
       "      <th>target</th>\n",
       "      <th>iter</th>\n",
       "      <th>train_mse</th>\n",
       "      <th>train_mape</th>\n",
       "      <th>train_r2</th>\n",
       "      <th>test_mse</th>\n",
       "      <th>test_mape</th>\n",
       "      <th>test_r2</th>\n",
       "      <th>independent_mse</th>\n",
       "      <th>independent_mape</th>\n",
       "      <th>independent_r2</th>\n",
       "      <th>runtime</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Mut</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>auc</td>\n",
       "      <td>0</td>\n",
       "      <td>1980.252745</td>\n",
       "      <td>1.210217e+08</td>\n",
       "      <td>0.649407</td>\n",
       "      <td>5223.408409</td>\n",
       "      <td>0.470750</td>\n",
       "      <td>-0.264799</td>\n",
       "      <td>9330.721311</td>\n",
       "      <td>3.689016e+08</td>\n",
       "      <td>-0.381141</td>\n",
       "      <td>0.168936</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Mut</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>auc</td>\n",
       "      <td>1</td>\n",
       "      <td>1939.922363</td>\n",
       "      <td>1.200003e+08</td>\n",
       "      <td>0.640354</td>\n",
       "      <td>7275.085786</td>\n",
       "      <td>0.505288</td>\n",
       "      <td>-0.246901</td>\n",
       "      <td>8794.825286</td>\n",
       "      <td>2.822469e+08</td>\n",
       "      <td>-0.301817</td>\n",
       "      <td>0.070691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Mut</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>auc</td>\n",
       "      <td>2</td>\n",
       "      <td>2044.156139</td>\n",
       "      <td>1.178403e+08</td>\n",
       "      <td>0.635741</td>\n",
       "      <td>6472.126192</td>\n",
       "      <td>1.040294</td>\n",
       "      <td>-0.512774</td>\n",
       "      <td>8941.883911</td>\n",
       "      <td>3.171360e+08</td>\n",
       "      <td>-0.323585</td>\n",
       "      <td>0.043169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Mut</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>auc</td>\n",
       "      <td>3</td>\n",
       "      <td>1986.790757</td>\n",
       "      <td>9.751971e+07</td>\n",
       "      <td>0.652618</td>\n",
       "      <td>5646.284841</td>\n",
       "      <td>0.414641</td>\n",
       "      <td>-0.643742</td>\n",
       "      <td>9526.372261</td>\n",
       "      <td>3.676249e+08</td>\n",
       "      <td>-0.410101</td>\n",
       "      <td>0.040309</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Mut</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>auc</td>\n",
       "      <td>4</td>\n",
       "      <td>2001.191339</td>\n",
       "      <td>1.128230e+08</td>\n",
       "      <td>0.646743</td>\n",
       "      <td>6571.044965</td>\n",
       "      <td>0.623259</td>\n",
       "      <td>-0.560188</td>\n",
       "      <td>8914.041797</td>\n",
       "      <td>4.004617e+08</td>\n",
       "      <td>-0.319464</td>\n",
       "      <td>0.048337</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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>235</th>\n",
       "      <td>Mut+RelativeExpr</td>\n",
       "      <td>KernelRidge</td>\n",
       "      <td>ic50</td>\n",
       "      <td>5</td>\n",
       "      <td>8.196746</td>\n",
       "      <td>2.230968e+01</td>\n",
       "      <td>0.616485</td>\n",
       "      <td>14.717264</td>\n",
       "      <td>16.094121</td>\n",
       "      <td>0.316389</td>\n",
       "      <td>10.795857</td>\n",
       "      <td>4.323257e+01</td>\n",
       "      <td>0.441244</td>\n",
       "      <td>0.082189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>236</th>\n",
       "      <td>Mut+RelativeExpr</td>\n",
       "      <td>KernelRidge</td>\n",
       "      <td>ic50</td>\n",
       "      <td>6</td>\n",
       "      <td>8.825907</td>\n",
       "      <td>2.382497e+01</td>\n",
       "      <td>0.595680</td>\n",
       "      <td>8.127384</td>\n",
       "      <td>25.845404</td>\n",
       "      <td>0.395887</td>\n",
       "      <td>11.089220</td>\n",
       "      <td>4.736885e+01</td>\n",
       "      <td>0.426061</td>\n",
       "      <td>0.092425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>237</th>\n",
       "      <td>Mut+RelativeExpr</td>\n",
       "      <td>KernelRidge</td>\n",
       "      <td>ic50</td>\n",
       "      <td>7</td>\n",
       "      <td>8.282871</td>\n",
       "      <td>2.230021e+01</td>\n",
       "      <td>0.612049</td>\n",
       "      <td>14.029435</td>\n",
       "      <td>14.134227</td>\n",
       "      <td>0.360051</td>\n",
       "      <td>10.745687</td>\n",
       "      <td>4.457547e+01</td>\n",
       "      <td>0.443841</td>\n",
       "      <td>0.089852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>238</th>\n",
       "      <td>Mut+RelativeExpr</td>\n",
       "      <td>KernelRidge</td>\n",
       "      <td>ic50</td>\n",
       "      <td>8</td>\n",
       "      <td>8.604582</td>\n",
       "      <td>2.161688e+01</td>\n",
       "      <td>0.599398</td>\n",
       "      <td>10.058312</td>\n",
       "      <td>22.571185</td>\n",
       "      <td>0.526447</td>\n",
       "      <td>10.801636</td>\n",
       "      <td>4.344088e+01</td>\n",
       "      <td>0.440945</td>\n",
       "      <td>0.084736</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>239</th>\n",
       "      <td>Mut+RelativeExpr</td>\n",
       "      <td>KernelRidge</td>\n",
       "      <td>ic50</td>\n",
       "      <td>9</td>\n",
       "      <td>8.445913</td>\n",
       "      <td>2.097890e+01</td>\n",
       "      <td>0.602123</td>\n",
       "      <td>13.264656</td>\n",
       "      <td>25.136422</td>\n",
       "      <td>0.431790</td>\n",
       "      <td>10.637797</td>\n",
       "      <td>4.226648e+01</td>\n",
       "      <td>0.449425</td>\n",
       "      <td>0.096710</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>240 rows × 14 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          feature_set       method target  iter    train_mse    train_mape  \\\n",
       "0                 Mut      XGBoost    auc     0  1980.252745  1.210217e+08   \n",
       "1                 Mut      XGBoost    auc     1  1939.922363  1.200003e+08   \n",
       "2                 Mut      XGBoost    auc     2  2044.156139  1.178403e+08   \n",
       "3                 Mut      XGBoost    auc     3  1986.790757  9.751971e+07   \n",
       "4                 Mut      XGBoost    auc     4  2001.191339  1.128230e+08   \n",
       "..                ...          ...    ...   ...          ...           ...   \n",
       "235  Mut+RelativeExpr  KernelRidge   ic50     5     8.196746  2.230968e+01   \n",
       "236  Mut+RelativeExpr  KernelRidge   ic50     6     8.825907  2.382497e+01   \n",
       "237  Mut+RelativeExpr  KernelRidge   ic50     7     8.282871  2.230021e+01   \n",
       "238  Mut+RelativeExpr  KernelRidge   ic50     8     8.604582  2.161688e+01   \n",
       "239  Mut+RelativeExpr  KernelRidge   ic50     9     8.445913  2.097890e+01   \n",
       "\n",
       "     train_r2     test_mse  test_mape   test_r2  independent_mse  \\\n",
       "0    0.649407  5223.408409   0.470750 -0.264799      9330.721311   \n",
       "1    0.640354  7275.085786   0.505288 -0.246901      8794.825286   \n",
       "2    0.635741  6472.126192   1.040294 -0.512774      8941.883911   \n",
       "3    0.652618  5646.284841   0.414641 -0.643742      9526.372261   \n",
       "4    0.646743  6571.044965   0.623259 -0.560188      8914.041797   \n",
       "..        ...          ...        ...       ...              ...   \n",
       "235  0.616485    14.717264  16.094121  0.316389        10.795857   \n",
       "236  0.595680     8.127384  25.845404  0.395887        11.089220   \n",
       "237  0.612049    14.029435  14.134227  0.360051        10.745687   \n",
       "238  0.599398    10.058312  22.571185  0.526447        10.801636   \n",
       "239  0.602123    13.264656  25.136422  0.431790        10.637797   \n",
       "\n",
       "     independent_mape  independent_r2   runtime  \n",
       "0        3.689016e+08       -0.381141  0.168936  \n",
       "1        2.822469e+08       -0.301817  0.070691  \n",
       "2        3.171360e+08       -0.323585  0.043169  \n",
       "3        3.676249e+08       -0.410101  0.040309  \n",
       "4        4.004617e+08       -0.319464  0.048337  \n",
       "..                ...             ...       ...  \n",
       "235      4.323257e+01        0.441244  0.082189  \n",
       "236      4.736885e+01        0.426061  0.092425  \n",
       "237      4.457547e+01        0.443841  0.089852  \n",
       "238      4.344088e+01        0.440945  0.084736  \n",
       "239      4.226648e+01        0.449425  0.096710  \n",
       "\n",
       "[240 rows x 14 columns]"
      ]
     },
     "execution_count": 144,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_table.to_csv('venetoclax_regression_results_all.csv', index=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'R^2 for AUC on the independent test set')"
      ]
     },
     "execution_count": 160,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(data=data_table[(data_table.target=='auc')],\n",
    "            x='feature_set', y='independent_r2', hue='method')\n",
    "plt.ylim(0, 1)\n",
    "plt.xlabel('Feature set')\n",
    "plt.ylabel('R^2 for predicted vs actual AUC')\n",
    "plt.title('R^2 for AUC on the independent test set')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'R^2 for IC50 on the independent test set')"
      ]
     },
     "execution_count": 161,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(data=data_table[(data_table.target=='ic50')],\n",
    "            x='feature_set', y='independent_r2', hue='method')\n",
    "plt.ylim(0, 1)\n",
    "plt.xlabel('Feature set')\n",
    "plt.ylabel('R^2 for predicted vs actual IC50')\n",
    "plt.title('R^2 for IC50 on the independent test set')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "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": "3.10.6"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {},
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}