R7_model_debugging.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 7️⃣ Error Analysis\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load libraries\n",
    "import os\n",
    "import warnings\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "from IPython.display import display\n",
    "import pickle\n",
    "import statsmodels.api as sm\n",
    "import statsmodels.stats.diagnostic as ss\n",
    "import scipy.stats as stats\n",
    "\n",
    "# Import helper function made ad-hoc\n",
    "# from R2_functions import univariate_outlier_removal\n",
    "from R3_functions import chain_snap\n",
    "from R4_functions import regression_report\n",
    "from R7_functions import plot_residuals\n",
    "\n",
    "from sklearn.model_selection import LearningCurveDisplay, ShuffleSplit, learning_curve\n",
    "\n",
    "# Set random seed\n",
    "np.random.seed(42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**1.** Load the train and validation datasets used in section 5: \n",
    "\n",
    "Create a baseline model and carry out a data cleaning procedure based on the previous findings; "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "DATA_TRAIN_PATH = os.path.join(\"data\", \"data_train_processed.parquet\")\n",
    "DATA_VAL_PATH = os.path.join(\"data\", \"data_val_processed.parquet\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading train data\n",
      "Loading validation data\n"
     ]
    }
   ],
   "source": [
    "data_dict = {}\n",
    "\n",
    "for path in [DATA_TRAIN_PATH, DATA_VAL_PATH]:\n",
    "    if not os.path.exists(path):\n",
    "        raise FileNotFoundError(f\"File {path} not found\")\n",
    "\n",
    "    # Get key to store data in the dictionary and tailor the message\n",
    "    key = \"train\" if \"train\" in path else \"validation\"\n",
    "\n",
    "    print(f\"Loading {key} data\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading train data\n",
      "Count NaNs post processing: 0\n",
      "Size dataframe: (543, 15)\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>Price</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>log(Price)</th>\n",
       "      <th>log(Area)</th>\n",
       "      <th>log(Room)</th>\n",
       "      <th>Zip_num</th>\n",
       "      <th>Region</th>\n",
       "      <th>Location</th>\n",
       "      <th>District</th>\n",
       "      <th>AR-ratio</th>\n",
       "      <th>log(AR-ratio)</th>\n",
       "      <th>Euclidean_distance</th>\n",
       "      <th>Haversine_distance</th>\n",
       "      <th>Manhattan_distance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>635000.0</td>\n",
       "      <td>81.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>13.36138</td>\n",
       "      <td>4.394449</td>\n",
       "      <td>1.098612</td>\n",
       "      <td>1078</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>27.0</td>\n",
       "      <td>3.295837</td>\n",
       "      <td>3.317122</td>\n",
       "      <td>3.346089</td>\n",
       "      <td>0.031533</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>400000.0</td>\n",
       "      <td>61.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>12.89922</td>\n",
       "      <td>4.110874</td>\n",
       "      <td>0.693147</td>\n",
       "      <td>1057</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>30.5</td>\n",
       "      <td>3.417727</td>\n",
       "      <td>1.947822</td>\n",
       "      <td>2.348155</td>\n",
       "      <td>0.036792</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Price  Area  Room  log(Price)  log(Area)  log(Room)  Zip_num  Region  \\\n",
       "0  635000.0  81.0   3.0    13.36138   4.394449   1.098612     1078       0   \n",
       "1  400000.0  61.0   2.0    12.89922   4.110874   0.693147     1057       0   \n",
       "\n",
       "   Location  District  AR-ratio  log(AR-ratio)  Euclidean_distance  \\\n",
       "0         7         8      27.0       3.295837            3.317122   \n",
       "1         5         7      30.5       3.417727            1.947822   \n",
       "\n",
       "   Haversine_distance  Manhattan_distance  \n",
       "0            3.346089            0.031533  \n",
       "1            2.348155            0.036792  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train data head: None\n",
      "Loading validation data\n",
      "Count NaNs post processing: 0\n",
      "Size dataframe: (181, 15)\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>Price</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>log(Price)</th>\n",
       "      <th>log(Area)</th>\n",
       "      <th>log(Room)</th>\n",
       "      <th>Zip_num</th>\n",
       "      <th>Region</th>\n",
       "      <th>Location</th>\n",
       "      <th>District</th>\n",
       "      <th>AR-ratio</th>\n",
       "      <th>log(AR-ratio)</th>\n",
       "      <th>Euclidean_distance</th>\n",
       "      <th>Haversine_distance</th>\n",
       "      <th>Manhattan_distance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>425000.0</td>\n",
       "      <td>68</td>\n",
       "      <td>3</td>\n",
       "      <td>12.959844</td>\n",
       "      <td>4.219508</td>\n",
       "      <td>1.098612</td>\n",
       "      <td>1098</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>22.666667</td>\n",
       "      <td>3.120895</td>\n",
       "      <td>3.606198</td>\n",
       "      <td>4.174069</td>\n",
       "      <td>0.071974</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>285000.0</td>\n",
       "      <td>96</td>\n",
       "      <td>4</td>\n",
       "      <td>12.560244</td>\n",
       "      <td>4.564348</td>\n",
       "      <td>1.386294</td>\n",
       "      <td>1104</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>3.178054</td>\n",
       "      <td>7.350449</td>\n",
       "      <td>8.209335</td>\n",
       "      <td>0.135124</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Price  Area  Room  log(Price)  log(Area)  log(Room)  Zip_num  Region  \\\n",
       "0  425000.0    68     3   12.959844   4.219508   1.098612     1098       0   \n",
       "1  285000.0    96     4   12.560244   4.564348   1.386294     1104       1   \n",
       "\n",
       "   Location  District   AR-ratio  log(AR-ratio)  Euclidean_distance  \\\n",
       "0         9         8  22.666667       3.120895            3.606198   \n",
       "1         0         4  24.000000       3.178054            7.350449   \n",
       "\n",
       "   Haversine_distance  Manhattan_distance  \n",
       "0            4.174069            0.071974  \n",
       "1            8.209335            0.135124  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "validation data head: None\n"
     ]
    }
   ],
   "source": [
    "# Load both train and validation data, apply the same processing,\n",
    "# store them in a dictionary for later use\n",
    "data_dict = {}\n",
    "\n",
    "for path in [DATA_TRAIN_PATH, DATA_VAL_PATH]:\n",
    "    if not os.path.exists(path):\n",
    "        raise FileNotFoundError(f\"File {path} not found\")\n",
    "\n",
    "    # Get key to store data in the dictionary and tailor the message\n",
    "    key = \"train\" if \"train\" in path else \"validation\"\n",
    "\n",
    "    print(f\"Loading {key} data\")\n",
    "    # Load data and process it\n",
    "    df = (\n",
    "        pd.read_parquet(path)\n",
    "        # drop Address and Zipcode as they are not useful\n",
    "        .drop(columns=[\"Address\", \"Zip\"])\n",
    "        # Drop Lat and Lon we will used the distance to the city center\n",
    "        .drop(columns=[\"Lat\", \"Lon\"])\n",
    "        # Drop Delivery Area as it contains too many categories\n",
    "        .drop(columns=[\"Delivery_area\"])\n",
    "        # Discard all the values Price > 2M\n",
    "        .loc[lambda x: x.Price < 2e6]\n",
    "        # double check nan values as sklearn does not like them\n",
    "        .pipe(\n",
    "            chain_snap,\n",
    "            fn=lambda x: x.isna().sum().sum(),\n",
    "            msg=\"Count NaNs post processing\",\n",
    "        )\n",
    "        .pipe(chain_snap, fn=lambda x: x.shape, msg=\"Size dataframe\")\n",
    "        .pipe(chain_snap, fn=lambda x: display(x.head(2)), msg=f\"{key} data head\")\n",
    "    )\n",
    "\n",
    "    # store data in a dictionary\n",
    "    data_dict[key] = df\n",
    "\n",
    "\n",
    "# unpack data\n",
    "df_train = data_dict[\"train\"]\n",
    "df_val = data_dict[\"validation\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**2.** Load the best estimator which has been stored as a pickle file. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "FEATURES_CANDIDATE = [\n",
    "    \"log(Area)\",\n",
    "    \"log(Room)\",\n",
    "    \"log(AR-ratio)\",\n",
    "    \"Manhattan_distance\",\n",
    "    \"Region\",\n",
    "    \"Location\",\n",
    "]\n",
    "TARGET = \"Price\"  # we don't consider the log of the price as we will compare with the original price"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['2024-03-30-15-35-49_Ridge', '.DS_Store', '2024-03-30-15-22-49_Ridge', '2024-01-11-08-37-12_baseline_model', '2024-01-15-18-42-23_Ridge']\n"
     ]
    }
   ],
   "source": [
    "# Get list of folder in artifacts\n",
    "print(os.listdir(\"artifacts\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-1 {color: black;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;preprocessor&#x27;,\n",
       "                 ColumnTransformer(transformers=[(&#x27;num&#x27;,\n",
       "                                                  Pipeline(steps=[(&#x27;num_imputer&#x27;,\n",
       "                                                                   SimpleImputer(strategy=&#x27;median&#x27;)),\n",
       "                                                                  (&#x27;scaler&#x27;,\n",
       "                                                                   StandardScaler())]),\n",
       "                                                  [&#x27;log(Area)&#x27;, &#x27;log(Room)&#x27;,\n",
       "                                                   &#x27;log(AR-ratio)&#x27;,\n",
       "                                                   &#x27;Manhattan_distance&#x27;]),\n",
       "                                                 (&#x27;nom&#x27;,\n",
       "                                                  Pipeline(steps=[(&#x27;nom_imputer&#x27;,\n",
       "                                                                   SimpleImputer(strategy=&#x27;most_frequent&#x27;)),\n",
       "                                                                  (&#x27;nominal_encoder&#x27;,\n",
       "                                                                   OneHotEncoder(drop=&#x27;first&#x27;))]),\n",
       "                                                  [&#x27;Region&#x27;, &#x27;Location&#x27;])])),\n",
       "                (&#x27;Ridge&#x27;, Ridge(alpha=0.2535151515151515))])</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" ><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;preprocessor&#x27;,\n",
       "                 ColumnTransformer(transformers=[(&#x27;num&#x27;,\n",
       "                                                  Pipeline(steps=[(&#x27;num_imputer&#x27;,\n",
       "                                                                   SimpleImputer(strategy=&#x27;median&#x27;)),\n",
       "                                                                  (&#x27;scaler&#x27;,\n",
       "                                                                   StandardScaler())]),\n",
       "                                                  [&#x27;log(Area)&#x27;, &#x27;log(Room)&#x27;,\n",
       "                                                   &#x27;log(AR-ratio)&#x27;,\n",
       "                                                   &#x27;Manhattan_distance&#x27;]),\n",
       "                                                 (&#x27;nom&#x27;,\n",
       "                                                  Pipeline(steps=[(&#x27;nom_imputer&#x27;,\n",
       "                                                                   SimpleImputer(strategy=&#x27;most_frequent&#x27;)),\n",
       "                                                                  (&#x27;nominal_encoder&#x27;,\n",
       "                                                                   OneHotEncoder(drop=&#x27;first&#x27;))]),\n",
       "                                                  [&#x27;Region&#x27;, &#x27;Location&#x27;])])),\n",
       "                (&#x27;Ridge&#x27;, Ridge(alpha=0.2535151515151515))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" ><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">preprocessor: ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[(&#x27;num&#x27;,\n",
       "                                 Pipeline(steps=[(&#x27;num_imputer&#x27;,\n",
       "                                                  SimpleImputer(strategy=&#x27;median&#x27;)),\n",
       "                                                 (&#x27;scaler&#x27;, StandardScaler())]),\n",
       "                                 [&#x27;log(Area)&#x27;, &#x27;log(Room)&#x27;, &#x27;log(AR-ratio)&#x27;,\n",
       "                                  &#x27;Manhattan_distance&#x27;]),\n",
       "                                (&#x27;nom&#x27;,\n",
       "                                 Pipeline(steps=[(&#x27;nom_imputer&#x27;,\n",
       "                                                  SimpleImputer(strategy=&#x27;most_frequent&#x27;)),\n",
       "                                                 (&#x27;nominal_encoder&#x27;,\n",
       "                                                  OneHotEncoder(drop=&#x27;first&#x27;))]),\n",
       "                                 [&#x27;Region&#x27;, &#x27;Location&#x27;])])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" ><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">num</label><div class=\"sk-toggleable__content\"><pre>[&#x27;log(Area)&#x27;, &#x27;log(Room)&#x27;, &#x27;log(AR-ratio)&#x27;, &#x27;Manhattan_distance&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-4\" type=\"checkbox\" ><label for=\"sk-estimator-id-4\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(strategy=&#x27;median&#x27;)</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-5\" type=\"checkbox\" ><label for=\"sk-estimator-id-5\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">StandardScaler</label><div class=\"sk-toggleable__content\"><pre>StandardScaler()</pre></div></div></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-6\" type=\"checkbox\" ><label for=\"sk-estimator-id-6\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">nom</label><div class=\"sk-toggleable__content\"><pre>[&#x27;Region&#x27;, &#x27;Location&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-7\" type=\"checkbox\" ><label for=\"sk-estimator-id-7\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(strategy=&#x27;most_frequent&#x27;)</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-8\" type=\"checkbox\" ><label for=\"sk-estimator-id-8\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">OneHotEncoder</label><div class=\"sk-toggleable__content\"><pre>OneHotEncoder(drop=&#x27;first&#x27;)</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-9\" type=\"checkbox\" ><label for=\"sk-estimator-id-9\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Ridge</label><div class=\"sk-toggleable__content\"><pre>Ridge(alpha=0.2535151515151515)</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('num',\n",
       "                                                  Pipeline(steps=[('num_imputer',\n",
       "                                                                   SimpleImputer(strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   StandardScaler())]),\n",
       "                                                  ['log(Area)', 'log(Room)',\n",
       "                                                   'log(AR-ratio)',\n",
       "                                                   'Manhattan_distance']),\n",
       "                                                 ('nom',\n",
       "                                                  Pipeline(steps=[('nom_imputer',\n",
       "                                                                   SimpleImputer(strategy='most_frequent')),\n",
       "                                                                  ('nominal_encoder',\n",
       "                                                                   OneHotEncoder(drop='first'))]),\n",
       "                                                  ['Region', 'Location'])])),\n",
       "                ('Ridge', Ridge(alpha=0.2535151515151515))])"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>metrics</th>\n",
       "      <th>in-sample</th>\n",
       "      <th>out-of-sample</th>\n",
       "      <th>Difference (%)</th>\n",
       "      <th>Description</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>model</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Ridge</th>\n",
       "      <td>r-squared</td>\n",
       "      <td>0.840</td>\n",
       "      <td>0.777</td>\n",
       "      <td>-8.108108</td>\n",
       "      <td>Coefficient of Determination</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ridge</th>\n",
       "      <td>explained variance</td>\n",
       "      <td>0.841</td>\n",
       "      <td>0.783</td>\n",
       "      <td>-7.407407</td>\n",
       "      <td>Explained Variance Score</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ridge</th>\n",
       "      <td>RMSE</td>\n",
       "      <td>127800.591</td>\n",
       "      <td>160521.179</td>\n",
       "      <td>20.383969</td>\n",
       "      <td>Root Mean Squared Error (RMSE)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ridge</th>\n",
       "      <td>MAE</td>\n",
       "      <td>76344.930</td>\n",
       "      <td>93887.806</td>\n",
       "      <td>18.684936</td>\n",
       "      <td>Mean Absolute Error (MAE)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ridge</th>\n",
       "      <td>max error</td>\n",
       "      <td>942098.599</td>\n",
       "      <td>1031360.528</td>\n",
       "      <td>8.654775</td>\n",
       "      <td>The maximum residual error</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  metrics   in-sample  out-of-sample  Difference (%)  \\\n",
       "model                                                                  \n",
       "Ridge           r-squared       0.840          0.777       -8.108108   \n",
       "Ridge  explained variance       0.841          0.783       -7.407407   \n",
       "Ridge                RMSE  127800.591     160521.179       20.383969   \n",
       "Ridge                 MAE   76344.930      93887.806       18.684936   \n",
       "Ridge           max error  942098.599    1031360.528        8.654775   \n",
       "\n",
       "                          Description  \n",
       "model                                  \n",
       "Ridge    Coefficient of Determination  \n",
       "Ridge        Explained Variance Score  \n",
       "Ridge  Root Mean Squared Error (RMSE)  \n",
       "Ridge       Mean Absolute Error (MAE)  \n",
       "Ridge      The maximum residual error  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Load Candidate model and its corresponding performance table\n",
    "CANDIDATE_PATH = os.path.join(\"artifacts\", \"2024-03-30-15-35-49_Ridge\")\n",
    "MODEL_PATH = os.path.join(CANDIDATE_PATH, \"Ridge.pkl\")\n",
    "candidate_model = pickle.load(open(MODEL_PATH, \"rb\"))\n",
    "display(candidate_model)\n",
    "\n",
    "# import model performance table\n",
    "REPORT_PATH = os.path.join(CANDIDATE_PATH, \"ridge_performance.csv\")\n",
    "df_performance_candidate = pd.read_csv(REPORT_PATH, index_col=0)\n",
    "display(df_performance_candidate)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Case: Regression problem**\n",
    "\n",
    "When dealing with a regression problem, one should analyse the residuals, which are the difference between the predicted and observed values of the dependent variable (target) and subsequently:\n",
    "\n",
    "**•** Plot the residuals versus the fitted values, and determine qualitatively any phenomenon of [heteroskedasticity](https://www.displayr.com/what-is-heteroscedasticity/#:~:text=The%20most%20widely%20used%20test,used%20in%20the%20original%20model.)\n",
    "\n",
    "🧑‍🔬 To get a quantitative assessment of heteroskedasticity use the [*Breusch-Pagan test*](https://www.statsmodels.org/stable/generated/statsmodels.stats.diagnostic.het_breuschpagan.html)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Computing residuals for train data\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>Price</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>log(Price)</th>\n",
       "      <th>log(Area)</th>\n",
       "      <th>log(Room)</th>\n",
       "      <th>Zip_num</th>\n",
       "      <th>Region</th>\n",
       "      <th>Location</th>\n",
       "      <th>District</th>\n",
       "      <th>AR-ratio</th>\n",
       "      <th>log(AR-ratio)</th>\n",
       "      <th>Euclidean_distance</th>\n",
       "      <th>Haversine_distance</th>\n",
       "      <th>Manhattan_distance</th>\n",
       "      <th>y_pred</th>\n",
       "      <th>residuals</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>635000.0</td>\n",
       "      <td>81.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>13.361</td>\n",
       "      <td>4.394</td>\n",
       "      <td>1.099</td>\n",
       "      <td>1078</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>27.000</td>\n",
       "      <td>3.296</td>\n",
       "      <td>3.317</td>\n",
       "      <td>3.346</td>\n",
       "      <td>0.032</td>\n",
       "      <td>657362.162</td>\n",
       "      <td>-22362.162</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>400000.0</td>\n",
       "      <td>61.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>12.899</td>\n",
       "      <td>4.111</td>\n",
       "      <td>0.693</td>\n",
       "      <td>1057</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>30.500</td>\n",
       "      <td>3.418</td>\n",
       "      <td>1.948</td>\n",
       "      <td>2.348</td>\n",
       "      <td>0.037</td>\n",
       "      <td>456822.940</td>\n",
       "      <td>-56822.940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>695000.0</td>\n",
       "      <td>88.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>13.452</td>\n",
       "      <td>4.477</td>\n",
       "      <td>1.099</td>\n",
       "      <td>1053</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>29.333</td>\n",
       "      <td>3.379</td>\n",
       "      <td>1.226</td>\n",
       "      <td>1.430</td>\n",
       "      <td>0.025</td>\n",
       "      <td>647048.593</td>\n",
       "      <td>47951.407</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1050000.0</td>\n",
       "      <td>119.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>13.864</td>\n",
       "      <td>4.779</td>\n",
       "      <td>1.609</td>\n",
       "      <td>1075</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>23.800</td>\n",
       "      <td>3.170</td>\n",
       "      <td>2.846</td>\n",
       "      <td>3.146</td>\n",
       "      <td>0.053</td>\n",
       "      <td>888404.921</td>\n",
       "      <td>161595.079</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>300000.0</td>\n",
       "      <td>54.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>12.612</td>\n",
       "      <td>3.989</td>\n",
       "      <td>0.693</td>\n",
       "      <td>1055</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>27.000</td>\n",
       "      <td>3.296</td>\n",
       "      <td>2.922</td>\n",
       "      <td>3.370</td>\n",
       "      <td>0.057</td>\n",
       "      <td>394687.897</td>\n",
       "      <td>-94687.897</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Price   Area  Room  log(Price)  log(Area)  log(Room)  Zip_num  Region  \\\n",
       "0   635000.0   81.0   3.0      13.361      4.394      1.099     1078       0   \n",
       "1   400000.0   61.0   2.0      12.899      4.111      0.693     1057       0   \n",
       "2   695000.0   88.0   3.0      13.452      4.477      1.099     1053       0   \n",
       "3  1050000.0  119.0   5.0      13.864      4.779      1.609     1075       0   \n",
       "4   300000.0   54.0   2.0      12.612      3.989      0.693     1055       0   \n",
       "\n",
       "   Location  District  AR-ratio  log(AR-ratio)  Euclidean_distance  \\\n",
       "0         7         8    27.000          3.296               3.317   \n",
       "1         5         7    30.500          3.418               1.948   \n",
       "2         5         3    29.333          3.379               1.226   \n",
       "3         7         5    23.800          3.170               2.846   \n",
       "4         5         5    27.000          3.296               2.922   \n",
       "\n",
       "   Haversine_distance  Manhattan_distance      y_pred   residuals  \n",
       "0               3.346               0.032  657362.162  -22362.162  \n",
       "1               2.348               0.037  456822.940  -56822.940  \n",
       "2               1.430               0.025  647048.593   47951.407  \n",
       "3               3.146               0.053  888404.921  161595.079  \n",
       "4               3.370               0.057  394687.897  -94687.897  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Computing residuals for validation data\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>Price</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>log(Price)</th>\n",
       "      <th>log(Area)</th>\n",
       "      <th>log(Room)</th>\n",
       "      <th>Zip_num</th>\n",
       "      <th>Region</th>\n",
       "      <th>Location</th>\n",
       "      <th>District</th>\n",
       "      <th>AR-ratio</th>\n",
       "      <th>log(AR-ratio)</th>\n",
       "      <th>Euclidean_distance</th>\n",
       "      <th>Haversine_distance</th>\n",
       "      <th>Manhattan_distance</th>\n",
       "      <th>y_pred</th>\n",
       "      <th>residuals</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>425000.0</td>\n",
       "      <td>68</td>\n",
       "      <td>3</td>\n",
       "      <td>12.960</td>\n",
       "      <td>4.220</td>\n",
       "      <td>1.099</td>\n",
       "      <td>1098</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>22.667</td>\n",
       "      <td>3.121</td>\n",
       "      <td>3.606</td>\n",
       "      <td>4.174</td>\n",
       "      <td>0.072</td>\n",
       "      <td>463212.431</td>\n",
       "      <td>-38212.431</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>285000.0</td>\n",
       "      <td>96</td>\n",
       "      <td>4</td>\n",
       "      <td>12.560</td>\n",
       "      <td>4.564</td>\n",
       "      <td>1.386</td>\n",
       "      <td>1104</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>24.000</td>\n",
       "      <td>3.178</td>\n",
       "      <td>7.350</td>\n",
       "      <td>8.209</td>\n",
       "      <td>0.135</td>\n",
       "      <td>350322.899</td>\n",
       "      <td>-65322.899</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>270000.0</td>\n",
       "      <td>84</td>\n",
       "      <td>3</td>\n",
       "      <td>12.506</td>\n",
       "      <td>4.431</td>\n",
       "      <td>1.099</td>\n",
       "      <td>1106</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>28.000</td>\n",
       "      <td>3.332</td>\n",
       "      <td>9.243</td>\n",
       "      <td>10.090</td>\n",
       "      <td>0.158</td>\n",
       "      <td>297404.777</td>\n",
       "      <td>-27404.777</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>450000.0</td>\n",
       "      <td>53</td>\n",
       "      <td>2</td>\n",
       "      <td>13.017</td>\n",
       "      <td>3.970</td>\n",
       "      <td>0.693</td>\n",
       "      <td>1051</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>26.500</td>\n",
       "      <td>3.277</td>\n",
       "      <td>1.442</td>\n",
       "      <td>1.590</td>\n",
       "      <td>0.027</td>\n",
       "      <td>411364.437</td>\n",
       "      <td>38635.563</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>549000.0</td>\n",
       "      <td>101</td>\n",
       "      <td>4</td>\n",
       "      <td>13.216</td>\n",
       "      <td>4.615</td>\n",
       "      <td>1.386</td>\n",
       "      <td>1031</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>25.250</td>\n",
       "      <td>3.229</td>\n",
       "      <td>1.922</td>\n",
       "      <td>2.005</td>\n",
       "      <td>0.030</td>\n",
       "      <td>542723.137</td>\n",
       "      <td>6276.863</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Price  Area  Room  log(Price)  log(Area)  log(Room)  Zip_num  Region  \\\n",
       "0  425000.0    68     3      12.960      4.220      1.099     1098       0   \n",
       "1  285000.0    96     4      12.560      4.564      1.386     1104       1   \n",
       "2  270000.0    84     3      12.506      4.431      1.099     1106       1   \n",
       "3  450000.0    53     2      13.017      3.970      0.693     1051       0   \n",
       "4  549000.0   101     4      13.216      4.615      1.386     1031       0   \n",
       "\n",
       "   Location  District  AR-ratio  log(AR-ratio)  Euclidean_distance  \\\n",
       "0         9         8    22.667          3.121               3.606   \n",
       "1         0         4    24.000          3.178               7.350   \n",
       "2         0         6    28.000          3.332               9.243   \n",
       "3         5         1    26.500          3.277               1.442   \n",
       "4         3         1    25.250          3.229               1.922   \n",
       "\n",
       "   Haversine_distance  Manhattan_distance      y_pred  residuals  \n",
       "0               4.174               0.072  463212.431 -38212.431  \n",
       "1               8.209               0.135  350322.899 -65322.899  \n",
       "2              10.090               0.158  297404.777 -27404.777  \n",
       "3               1.590               0.027  411364.437  38635.563  \n",
       "4               2.005               0.030  542723.137   6276.863  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# loop through dataframes\n",
    "for key, df in data_dict.items():\n",
    "    print(f\"Computing residuals for {key} data\")\n",
    "    # compute predictions\n",
    "    y_pred = np.exp(candidate_model.predict(df[FEATURES_CANDIDATE]))\n",
    "    # compute residuals\n",
    "    residuals = df[TARGET] - y_pred\n",
    "    # store predictions and residuals in dataframe\n",
    "    df[\"y_pred\"] = y_pred\n",
    "    df[\"residuals\"] = residuals\n",
    "    # print out the updated dfs with predictions and residuals\n",
    "    display(df.head().round(3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "warnings.filterwarnings(\"ignore\")\n",
    "from matplotlib import ticker \n",
    "\n",
    "fig, ax = plt.subplots(figsize=(12, 6))\n",
    "\n",
    "# Set line width and transparency\n",
    "lw = 2\n",
    "alpha = 0.75\n",
    "\n",
    "# Define colors for each dataset\n",
    "train_color = '#4C8BF5'\n",
    "validation_color = '#F9812A'\n",
    "\n",
    "# Loop through data and plot\n",
    "for key, df in data_dict.items():\n",
    "    if key == \"train\":\n",
    "        scatter_color = train_color\n",
    "        line_color = \"blue\"\n",
    "        label_residuals = \"Residual (Train set)\"\n",
    "        label_trend = \"Residual Trend (Train set)\"\n",
    "    elif key == \"validation\":\n",
    "        scatter_color = validation_color\n",
    "        line_color = \"orange\"\n",
    "        label_residuals = \"Residual (Validation set)\"\n",
    "        label_trend = \"Residual Trend (Validation set)\"\n",
    "    else:\n",
    "        raise ValueError(f\"Unknown key {key}\")\n",
    "\n",
    "    # Scatter plot\n",
    "    sns.scatterplot(\n",
    "        x=df[\"y_pred\"],\n",
    "        y=df[\"residuals\"],\n",
    "        color=scatter_color,\n",
    "        alpha=0.6,\n",
    "        label=label_residuals,\n",
    "        ax=ax,\n",
    "    )\n",
    "\n",
    "    # Generate trend lines\n",
    "    sns.regplot(\n",
    "        x=df[\"y_pred\"],\n",
    "        y=df[\"residuals\"],\n",
    "        scatter=False,\n",
    "        lowess=True,\n",
    "        line_kws={\"color\": line_color, \"lw\": lw, \"alpha\": alpha},\n",
    "        ax=ax,\n",
    "        label=label_trend,\n",
    "    )\n",
    "\n",
    "# Add horizontal line for reference at y=0\n",
    "ax.axhline(0, color=\"red\", linestyle=\"--\", linewidth=1.5)\n",
    "\n",
    "# Customize plot\n",
    "ax.set_title(\"Residuals vs. Fitted\", fontsize=16)\n",
    "ax.set_xlabel(\"Fitted values\", fontsize=14)\n",
    "ax.set_ylabel(\"Residuals\", fontsize=14)\n",
    "\n",
    "\n",
    "# Add legend\n",
    "ax.legend(loc=\"upper right\", frameon=True, framealpha=0.9)\n",
    "\n",
    "# Show plot\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "\n",
    "# Show the plot\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**•** Plot the distribution of the residuals and measure the mean and the residual standard error. Contextualize these values and assess if they are acceptable using business knowledge."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Residual Statistics (in k€):\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_ebb64\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_ebb64_level0_col0\" class=\"col_heading level0 col0\" >residuals_train</th>\n",
       "      <th id=\"T_ebb64_level0_col1\" class=\"col_heading level0 col1\" >residuals_validation</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_ebb64_level0_row0\" class=\"row_heading level0 row0\" >mean</th>\n",
       "      <td id=\"T_ebb64_row0_col0\" class=\"data row0 col0\" >11.6 k€</td>\n",
       "      <td id=\"T_ebb64_row0_col1\" class=\"data row0 col1\" >27.4 k€</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ebb64_level0_row1\" class=\"row_heading level0 row1\" >std</th>\n",
       "      <td id=\"T_ebb64_row1_col0\" class=\"data row1 col0\" >127.4 k€</td>\n",
       "      <td id=\"T_ebb64_row1_col1\" class=\"data row1 col1\" >158.6 k€</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1a19e4a30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Initialize an empty DataFrame to store results\n",
    "df_res = pd.DataFrame()\n",
    "\n",
    "# Loop through the df and compute the mean and standard deviation of residuals\n",
    "for key, df in data_dict.items():\n",
    "    # Scale residuals to thousands and compute mean and std\n",
    "    res = (\n",
    "        df.assign(residuals=lambda x: x[\"residuals\"] / 1000)  # Scale residuals\n",
    "        .agg({\"residuals\": [\"mean\", \"std\"]})  # Compute mean and std\n",
    "        .rename(columns={\"residuals\": f\"residuals_{key}\"})  # Rename columns for each set\n",
    "    )\n",
    "    \n",
    "    # Concatenate results\n",
    "    df_res = pd.concat([df_res, res], axis=1)\n",
    "\n",
    "print(\"Residual Statistics (in k€):\")\n",
    "# Format values as k€ and display\n",
    "display( df_res.style.format(\"{:,.1f} k€\"))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***Intrerpretation of the mean, residual standard error***\n",
    "- When the model is correctly specified, the mean value of residuals should be zero\n",
    "- *The **mean value** of the residual indicates whether the model yields a systematic overestimation (negative value) or underestimation (positive value) on its prediction.*\n",
    "- *The **residual standard error** is the average amount of actual values differing from the model's predictions.* \n",
    "\n",
    "For the candidate model, we would expect that we would estimate the price of a house with on average 28k€ less than the actual value and with 68% of the predicted price will be $\\pm$ 159k€ (note: 68% is because we assume the residual normally distributed hence 68% of the data is roughly one standard deviation from the average). "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*For example, for a hypothetical house with an Area of 100 $m^{2}$, 5 Rooms, 0.15 distant from the centre expressed in Manhattan_distance, located in Region 1 and location 0, the candidate model predicts*:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The estimated price for the sample is 354 k€ \n",
      "with 68% of chance of having the true value between 222 k€ and 539 k€.\n"
     ]
    }
   ],
   "source": [
    "# Get results based on example provided above\n",
    "sample = pd.DataFrame(\n",
    "    data={\n",
    "        \"log(Area)\": np.log(100),\n",
    "        \"log(Room)\": np.log(5),\n",
    "        \"log(AR-ratio)\": np.log(100 / 5),\n",
    "        \"Manhattan_distance\": 0.15,\n",
    "        \"Region\": 1,\n",
    "        \"Location\": 0,\n",
    "    },\n",
    "    index=[0],\n",
    ")\n",
    "\n",
    "# express prices in k€\n",
    "house_price_estimation = np.exp(candidate_model.predict(sample))[0] / 1e3\n",
    "mean_residual = df_val[\"residuals\"].mean() / 1e3\n",
    "std_residual = df_val[\"residuals\"].std() / 1e3\n",
    "\n",
    "print(\n",
    "    f\"\"\"The estimated price for the sample is {house_price_estimation:.0f} k€ \n",
    "with 68% of chance of having the true value between {house_price_estimation + mean_residual - std_residual:.0f} k€ and {house_price_estimation + mean_residual + std_residual:.0f} k€.\"\"\"\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*This example showed that such a prediction would not be useful, as the uncertainty range is too high.* "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Check high residual values, which indicate a poor prediction, and try to spot any pattern by analyzing the corresponding input values. Investigate the distribution of the model input (features) and apply statistical tests by comparing outliers versus non-outlier values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "o Show summary statistics in validation data\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>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>27.4</td>\n",
       "      <td>158.6</td>\n",
       "      <td>-293.1</td>\n",
       "      <td>-46.5</td>\n",
       "      <td>1.0</td>\n",
       "      <td>53.4</td>\n",
       "      <td>1031.4</td>\n",
       "      <td>181.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mean    std    min   25%  50%   75%     max  count\n",
       "0  27.4  158.6 -293.1 -46.5  1.0  53.4  1031.4  181.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Select which residuals to plot among 'train' and 'validation' sets \n",
    "key = \"validation\"  # options are 'train' or 'validation\n",
    "\n",
    "# focus only on the validation set\n",
    "df_plot = df_val.copy()\n",
    "plot_stast = df_plot[\"residuals\"].describe()\n",
    "\n",
    "# compute summary statistics to visualize on the plot\n",
    "s_stast = pd.concat(\n",
    "    [\n",
    "        df_plot[\"residuals\"]\n",
    "        # get summary statistics\n",
    "        .describe()\n",
    "        # divide by 1000 to get k€\n",
    "        .apply(lambda x: x / 1000)\n",
    "        # round to 1 decimal\n",
    "        .round(1)\n",
    "        # drop count as it is not useful\n",
    "        .drop(\"count\"),\n",
    "        # re-add count at the end of the series\n",
    "        pd.Series(df_plot[\"residuals\"].count(), index=[\"count\"]),\n",
    "    ]\n",
    ")\n",
    "\n",
    "print(f\"o Show summary statistics in {key} data\")\n",
    "display(s_stast.to_frame().T)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#PLOT RESIDUALS DISTRIBUTION in both as histogram and boxplot\n",
    "plot_residuals(data_dict, dataset_key=\"validation\")  # options are 'train' or 'validation'\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- The above visualisation showed that the residuals are not normally distrubuted, due to a heavy right skew. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "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>Price</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>Manhattan_distance</th>\n",
       "      <th>residuals</th>\n",
       "      <th>outliers</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>575.0</td>\n",
       "      <td>183</td>\n",
       "      <td>5</td>\n",
       "      <td>0.058278</td>\n",
       "      <td>-293.1</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>1099.0</td>\n",
       "      <td>191</td>\n",
       "      <td>6</td>\n",
       "      <td>0.044197</td>\n",
       "      <td>-270.0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>1095.0</td>\n",
       "      <td>204</td>\n",
       "      <td>6</td>\n",
       "      <td>0.022995</td>\n",
       "      <td>-269.3</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>184</th>\n",
       "      <td>799.0</td>\n",
       "      <td>146</td>\n",
       "      <td>5</td>\n",
       "      <td>0.063903</td>\n",
       "      <td>-241.0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>350.0</td>\n",
       "      <td>110</td>\n",
       "      <td>5</td>\n",
       "      <td>0.063693</td>\n",
       "      <td>-199.8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>163</th>\n",
       "      <td>915.0</td>\n",
       "      <td>88</td>\n",
       "      <td>5</td>\n",
       "      <td>0.010290</td>\n",
       "      <td>222.4</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>1495.0</td>\n",
       "      <td>175</td>\n",
       "      <td>6</td>\n",
       "      <td>0.010027</td>\n",
       "      <td>226.9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>400.0</td>\n",
       "      <td>25</td>\n",
       "      <td>2</td>\n",
       "      <td>0.079482</td>\n",
       "      <td>238.2</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>950.0</td>\n",
       "      <td>94</td>\n",
       "      <td>1</td>\n",
       "      <td>0.027149</td>\n",
       "      <td>248.8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>1698.0</td>\n",
       "      <td>205</td>\n",
       "      <td>5</td>\n",
       "      <td>0.053619</td>\n",
       "      <td>269.6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>84</th>\n",
       "      <td>950.0</td>\n",
       "      <td>174</td>\n",
       "      <td>6</td>\n",
       "      <td>0.123966</td>\n",
       "      <td>274.9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>1300.0</td>\n",
       "      <td>139</td>\n",
       "      <td>5</td>\n",
       "      <td>0.016075</td>\n",
       "      <td>277.3</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>950.0</td>\n",
       "      <td>133</td>\n",
       "      <td>7</td>\n",
       "      <td>0.253201</td>\n",
       "      <td>323.8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>899.0</td>\n",
       "      <td>128</td>\n",
       "      <td>5</td>\n",
       "      <td>0.104190</td>\n",
       "      <td>334.6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>1450.0</td>\n",
       "      <td>150</td>\n",
       "      <td>3</td>\n",
       "      <td>0.005574</td>\n",
       "      <td>339.1</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>178</th>\n",
       "      <td>1250.0</td>\n",
       "      <td>180</td>\n",
       "      <td>9</td>\n",
       "      <td>0.036860</td>\n",
       "      <td>354.3</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>1095.0</td>\n",
       "      <td>134</td>\n",
       "      <td>3</td>\n",
       "      <td>0.057440</td>\n",
       "      <td>356.7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>895.0</td>\n",
       "      <td>95</td>\n",
       "      <td>3</td>\n",
       "      <td>0.104190</td>\n",
       "      <td>462.3</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111</th>\n",
       "      <td>1775.0</td>\n",
       "      <td>177</td>\n",
       "      <td>9</td>\n",
       "      <td>0.047147</td>\n",
       "      <td>497.4</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>168</th>\n",
       "      <td>1599.0</td>\n",
       "      <td>139</td>\n",
       "      <td>5</td>\n",
       "      <td>0.035882</td>\n",
       "      <td>649.7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>1395.0</td>\n",
       "      <td>133</td>\n",
       "      <td>3</td>\n",
       "      <td>0.051598</td>\n",
       "      <td>733.1</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>1750.0</td>\n",
       "      <td>130</td>\n",
       "      <td>3</td>\n",
       "      <td>0.057651</td>\n",
       "      <td>1031.4</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Price  Area  Room  Manhattan_distance  residuals  outliers\n",
       "83    575.0   183     5            0.058278     -293.1      True\n",
       "107  1099.0   191     6            0.044197     -270.0      True\n",
       "158  1095.0   204     6            0.022995     -269.3      True\n",
       "184   799.0   146     5            0.063903     -241.0      True\n",
       "117   350.0   110     5            0.063693     -199.8      True\n",
       "163   915.0    88     5            0.010290      222.4      True\n",
       "149  1495.0   175     6            0.010027      226.9      True\n",
       "102   400.0    25     2            0.079482      238.2      True\n",
       "97    950.0    94     1            0.027149      248.8      True\n",
       "93   1698.0   205     5            0.053619      269.6      True\n",
       "84    950.0   174     6            0.123966      274.9      True\n",
       "110  1300.0   139     5            0.016075      277.3      True\n",
       "26    950.0   133     7            0.253201      323.8      True\n",
       "119   899.0   128     5            0.104190      334.6      True\n",
       "34   1450.0   150     3            0.005574      339.1      True\n",
       "178  1250.0   180     9            0.036860      354.3      True\n",
       "53   1095.0   134     3            0.057440      356.7      True\n",
       "71    895.0    95     3            0.104190      462.3      True\n",
       "111  1775.0   177     9            0.047147      497.4      True\n",
       "168  1599.0   139     5            0.035882      649.7      True\n",
       "121  1395.0   133     3            0.051598      733.1      True\n",
       "70   1750.0   130     3            0.057651     1031.4      True"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "• Price and Residuals are in k€\n",
      "• Number of samples in validation set = 181\n",
      "• Absolute number of outliers in validation set = 22\n",
      "• Percentage of outliers in validation set = 12.15%\n"
     ]
    }
   ],
   "source": [
    "# Let us create a new variable called `outliers` that flags `True` outliers based on the IQR method\n",
    "\n",
    "# Calculate IQR and thresholds for outliers\n",
    "q1 = df_val[\"residuals\"].quantile(0.25)\n",
    "q3 = df_val[\"residuals\"].quantile(0.75)\n",
    "iqr = q3 - q1\n",
    "lower_bound = q1 - 1.5 * iqr\n",
    "upper_bound = q3 + 1.5 * iqr\n",
    "\n",
    "# Create a DataFrame for error analysis\n",
    "df_error_analysis = (\n",
    "    df_val\n",
    "    .assign(\n",
    "        # Identify outliers based on residual thresholds\n",
    "        outliers=lambda x: (x[\"residuals\"] < lower_bound) | (x[\"residuals\"] > upper_bound),\n",
    "        # Convert Price and residuals to thousands of euros and round residuals\n",
    "        Price=lambda x: x[\"Price\"] / 1000,\n",
    "        residuals=lambda x: round(x[\"residuals\"] / 1000, 1)\n",
    "    )\n",
    "    # Select only the relevant columns\n",
    "    .loc[:, [\"Price\", \"Area\", \"Room\", \"Manhattan_distance\", \"residuals\", \"outliers\"]]\n",
    "    .sort_values(by=\"residuals\", ascending=True)\n",
    ")\n",
    "\n",
    "# Display outliers\n",
    "display(df_error_analysis.loc[df_error_analysis[\"outliers\"]])\n",
    "\n",
    "# Store summary information\n",
    "num_samples = len(df_error_analysis)\n",
    "num_outliers = df_error_analysis[\"outliers\"].sum()\n",
    "percentage_outliers = (num_outliers / num_samples) * 100\n",
    "\n",
    "print(\"• Price and Residuals are in k€\")\n",
    "print(f\"• Number of samples in validation set = {num_samples}\")\n",
    "print(f\"• Absolute number of outliers in validation set = {num_outliers}\")\n",
    "print(f\"• Percentage of outliers in validation set = {percentage_outliers:.2f}%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x1000 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Define custom palette for outliers\n",
    "palette = [\"#4C8BF5\", \"#F9812A\"]  # Blue for non-outliers, Orange for outliers\n",
    "\n",
    "# Create subplots\n",
    "fig, axs = plt.subplots(nrows=2, ncols=2, figsize=(12, 10))\n",
    "\n",
    "# 1. Histogram of Price grouped by outliers\n",
    "sns.histplot(\n",
    "    data=df_error_analysis,\n",
    "    x=\"Price\",\n",
    "    hue=\"outliers\",\n",
    "    multiple=\"stack\",\n",
    "    palette=palette,\n",
    "    ax=axs[0, 0]\n",
    ")\n",
    "axs[0, 0].set_title(\"Price Distribution by Outliers\")\n",
    "\n",
    "# 2. Boxplot and swarmplot of Area grouped by outliers\n",
    "sns.boxplot(\n",
    "    data=df_error_analysis,\n",
    "    y=\"Area\",\n",
    "    x=\"outliers\",\n",
    "    palette=palette,\n",
    "    ax=axs[0, 1],\n",
    "    boxprops=dict(alpha=0.5)\n",
    ")\n",
    "sns.swarmplot(\n",
    "    data=df_error_analysis,\n",
    "    y=\"Area\",\n",
    "    x=\"outliers\",\n",
    "    color=\"black\",\n",
    "    size=4,\n",
    "    alpha=0.5,\n",
    "    ax=axs[0, 1]\n",
    ")\n",
    "axs[0, 1].set_title(\"Area Distribution by Outliers\")\n",
    "\n",
    "# 3. Histogram of Area vs Room grouped by outliers\n",
    "sns.histplot(\n",
    "    data=df_error_analysis,\n",
    "    x=\"Area\",\n",
    "    y=\"Room\",\n",
    "    hue=\"outliers\",\n",
    "    palette=palette,\n",
    "    ax=axs[1, 0]\n",
    ")\n",
    "axs[1, 0].set_title(\"Area vs Room Distribution by Outliers\")\n",
    "\n",
    "# 4. Boxplot and swarmplot of Manhattan_distance grouped by outliers\n",
    "sns.boxplot(\n",
    "    data=df_error_analysis,\n",
    "    y=\"Manhattan_distance\",\n",
    "    x=\"outliers\",\n",
    "    palette=palette,\n",
    "    ax=axs[1, 1],\n",
    "    boxprops=dict(alpha=0.5)\n",
    ")\n",
    "sns.swarmplot(\n",
    "    data=df_error_analysis,\n",
    "    y=\"Manhattan_distance\",\n",
    "    x=\"outliers\",\n",
    "    color=\"black\",\n",
    "    size=4,\n",
    "    alpha=0.5,\n",
    "    ax=axs[1, 1]\n",
    ")\n",
    "axs[1, 1].set_title(\"Manhattan Distance by Outliers\")\n",
    "\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "• Summary statistics of variables which appear driving the outliers\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Price</th>\n",
       "      <td>22.0</td>\n",
       "      <td>1122.0</td>\n",
       "      <td>408.5</td>\n",
       "      <td>350.0</td>\n",
       "      <td>903.0</td>\n",
       "      <td>1095.0</td>\n",
       "      <td>1436.2</td>\n",
       "      <td>1775.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Area</th>\n",
       "      <td>22.0</td>\n",
       "      <td>142.4</td>\n",
       "      <td>43.4</td>\n",
       "      <td>25.0</td>\n",
       "      <td>128.5</td>\n",
       "      <td>139.0</td>\n",
       "      <td>176.5</td>\n",
       "      <td>205.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Room</th>\n",
       "      <td>22.0</td>\n",
       "      <td>4.9</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       count    mean    std    min    25%     50%     75%     max\n",
       "Price   22.0  1122.0  408.5  350.0  903.0  1095.0  1436.2  1775.0\n",
       "Area    22.0   142.4   43.4   25.0  128.5   139.0   176.5   205.0\n",
       "Room    22.0     4.9    2.0    1.0    3.0     5.0     6.0     9.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"• Summary statistics of variables which appear driving the outliers\")\n",
    "display(df_error_analysis.loc[df_error_analysis[\"outliers\"] == True, [\"Price\", \"Area\", \"Room\"]].describe().round(1).T)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<!-- **Final assessment**\n",
    "\n",
    "**•** Summarize the findings and, if necessary, retrain a new model based on the new knowledge from the error analysis. -->\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Summary Error Analysis**\n",
    "\n",
    "Based on the analysis carried out above, we found that: \n",
    "- 12.15% of samples in the validation  (i.e.,22 out of 181 ) are flagged as outliers based on the IQR method: \n",
    "    - 75% of the outliers having Price above 903k€ \n",
    "    - Outliers have an average Area = 142.4 $m^{2}$.\n",
    "    - The distance from the center does not influence outliers.\n",
    "- A high price tends to have a higher chance of unacceptable prediction error. The same can be applied for Area as it strongly correlates with Price.\n",
    "    - Most outliers start from a price of 903 K€ and 128.5 $m^{2}$, the 25% quantile of the Price and Area respectively. \n",
    "- Room=1 and Room=9 are very few samples, flagged outliers in all cases. We also found before that Room strongly correlates with Area and Price.\n",
    "\n",
    "- One sample has Room = 1 with a Price of 950k€ and 94 $m^{2}$; this example is rather peculiar as Dutch houses do generally not have such a high square meter per room. The same consideration can be made for two other samples with Room=9. These samples are expected to be extremely difficult to estimate with the current data. Furthermore, a significant amount is considered as an outlier when Room=5; hence, one can consider restricting the input range of **2 <= Room <= 4** to have reasonable accuracy.  \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Given the current findings, we may want to consider the impact of outliers affecting the accuracy of the model, and one should consider re-train the model after removing them. \n",
    "\n",
    "Let's analyze the current model performance separately for:\n",
    "- Expected range with **good** model performance: $Area \\leq 128.5 m^2$ AND $2 \\leq Room \\leq 4$\n",
    "- Expected range with **poor** model performance: *else*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Candidate Model performance for all samples in validation set:\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>Score</th>\n",
       "      <th>Description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>r-squared</th>\n",
       "      <td>0.777</td>\n",
       "      <td>Coefficient of Determination</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>explained variance</th>\n",
       "      <td>0.783</td>\n",
       "      <td>Explained Variance Score</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>160521.179</td>\n",
       "      <td>Root Mean Squared Error (RMSE)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>93887.806</td>\n",
       "      <td>Mean Absolute Error (MAE)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max error</th>\n",
       "      <td>1031360.528</td>\n",
       "      <td>The maximum residual error</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          Score                     Description\n",
       "r-squared                 0.777    Coefficient of Determination\n",
       "explained variance        0.783        Explained Variance Score\n",
       "RMSE                 160521.179  Root Mean Squared Error (RMSE)\n",
       "MAE                   93887.806       Mean Absolute Error (MAE)\n",
       "max error           1031360.528      The maximum residual error"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Candidate Model performance with restricted range in validation set [expected to be better]:\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>Score</th>\n",
       "      <th>Description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>r-squared</th>\n",
       "      <td>0.782</td>\n",
       "      <td>Coefficient of Determination</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>explained variance</th>\n",
       "      <td>0.783</td>\n",
       "      <td>Explained Variance Score</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>79906.248</td>\n",
       "      <td>Root Mean Squared Error (RMSE)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>55641.344</td>\n",
       "      <td>Mean Absolute Error (MAE)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max error</th>\n",
       "      <td>462275.345</td>\n",
       "      <td>The maximum residual error</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         Score                     Description\n",
       "r-squared                0.782    Coefficient of Determination\n",
       "explained variance       0.783        Explained Variance Score\n",
       "RMSE                 79906.248  Root Mean Squared Error (RMSE)\n",
       "MAE                  55641.344       Mean Absolute Error (MAE)\n",
       "max error           462275.345      The maximum residual error"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "• Number of samples in good model performance range = 131\n",
      "\n",
      "Candidate Model performance with restricted range in validation set [expected to be worse]:\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>Score</th>\n",
       "      <th>Description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>r-squared</th>\n",
       "      <td>0.564</td>\n",
       "      <td>Coefficient of Determination</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>explained variance</th>\n",
       "      <td>0.635</td>\n",
       "      <td>Explained Variance Score</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>276673.082</td>\n",
       "      <td>Root Mean Squared Error (RMSE)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>194093.538</td>\n",
       "      <td>Mean Absolute Error (MAE)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max error</th>\n",
       "      <td>1031360.528</td>\n",
       "      <td>The maximum residual error</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          Score                     Description\n",
       "r-squared                 0.564    Coefficient of Determination\n",
       "explained variance        0.635        Explained Variance Score\n",
       "RMSE                 276673.082  Root Mean Squared Error (RMSE)\n",
       "MAE                  194093.538       Mean Absolute Error (MAE)\n",
       "max error           1031360.528      The maximum residual error"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "• Number of samples in bad model performance range = 50\n"
     ]
    }
   ],
   "source": [
    "# Candidate Model Performance for All Samples\n",
    "print(\"Candidate Model performance for all samples in validation set:\")\n",
    "full_performance = regression_report(y_true=df_val[\"Price\"], y_pred=df_val[\"y_pred\"])\n",
    "display(full_performance)\n",
    "\n",
    "# Filter Data: Good Range\n",
    "df_good_range = df_val.loc[\n",
    "    lambda x: (x[\"Area\"] <= 128) & (x[\"Room\"] >= 2) & (x[\"Room\"] <= 4)\n",
    "]\n",
    "\n",
    "\n",
    "print(\"\\nCandidate Model performance with restricted range in validation set [expected to be better]:\")\n",
    "good_range_performance = regression_report(\n",
    "    y_true=df_good_range[\"Price\"], \n",
    "    y_pred=df_good_range[\"y_pred\"]\n",
    ")\n",
    "display(good_range_performance)\n",
    "print(f\"• Number of samples in good model performance range = {len(df_good_range)}\")\n",
    "\n",
    "# Filter Data: Bad Range\n",
    "df_bad_range = df_val.loc[~df_val.index.isin(df_good_range.index)]\n",
    "\n",
    "# Ensure the lengths of the datasets match\n",
    "assert len(df_val) == len(df_good_range) + len(df_bad_range), \"Dataframe length mismatch!\"\n",
    "\n",
    "print(\"\\nCandidate Model performance with restricted range in validation set [expected to be worse]:\")\n",
    "bad_range_performance = regression_report(\n",
    "    y_true=df_bad_range[\"Price\"], \n",
    "    y_pred=df_bad_range[\"y_pred\"]\n",
    ")\n",
    "display(bad_range_performance)\n",
    "print(f\"• Number of samples in bad model performance range = {len(df_bad_range)}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<!-- To minimize the prediction error across the entire price range up to 2M€ one can attempt to train two models:\n",
    "- one model which is trained on a subset of the feature range, namely **Area <= 128 & 2 <= Room <= 4**.\n",
    "- a second model with a **complementary range or Room > 4**.\n",
    "\n",
    "Combine both models via a look-up table with *Area* and *Room* as selectors. -->"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Generate the [learning curve](https://machinelearningmastery.com/learning-curves-for-diagnosing-machine-learning-model-performance/) to graphically represents the model's learning performance changes as the number of samples increases and assess if the model is underfitting, overfitting, or well-fitting .   \n",
    "\n",
    "    - Learning curves can be used to diagnose if datasets are not representative of the problem domain.\n",
    "\n",
    "- Finally, use this plot to evaluate whether collecting additional data can increare the model performance.\n",
    "\n",
    "    - Collecting additional data can sometimes improve the performance of a machine learning model. However, there are cases where adding more data may not necessarily lead to better performance.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 266, in __call__\n",
      "    return self._score(partial(_cached_call, None), estimator, X, y_true, **_kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 353, in _score\n",
      "    y_pred = method_caller(estimator, \"predict\", X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 86, in _cached_call\n",
      "    result, _ = _get_response_values(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_response.py\", line 109, in _get_response_values\n",
      "    y_pred, pos_label = estimator.predict(X), None\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 507, in predict\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [8] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 266, in __call__\n",
      "    return self._score(partial(_cached_call, None), estimator, X, y_true, **_kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 353, in _score\n",
      "    y_pred = method_caller(estimator, \"predict\", X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 86, in _cached_call\n",
      "    result, _ = _get_response_values(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_response.py\", line 109, in _get_response_values\n",
      "    y_pred, pos_label = estimator.predict(X), None\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 507, in predict\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [8] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 266, in __call__\n",
      "    return self._score(partial(_cached_call, None), estimator, X, y_true, **_kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 353, in _score\n",
      "    y_pred = method_caller(estimator, \"predict\", X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 86, in _cached_call\n",
      "    result, _ = _get_response_values(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_response.py\", line 109, in _get_response_values\n",
      "    y_pred, pos_label = estimator.predict(X), None\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 507, in predict\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [8] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 266, in __call__\n",
      "    return self._score(partial(_cached_call, None), estimator, X, y_true, **_kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 353, in _score\n",
      "    y_pred = method_caller(estimator, \"predict\", X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 86, in _cached_call\n",
      "    result, _ = _get_response_values(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_response.py\", line 109, in _get_response_values\n",
      "    y_pred, pos_label = estimator.predict(X), None\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 507, in predict\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [2] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "warnings.filterwarnings(\"ignore\")\n",
    "# Define learning curve parameters\n",
    "learning_curve_params = {\n",
    "    \"X\": df_train[FEATURES_CANDIDATE],          # provide the features\n",
    "    \"y\": df_train[\"log(Price)\"],                # provide the target\n",
    "    \"train_sizes\": np.linspace(0.2, 1.0, 10),   # provide the train sizes (20% to 100%)\n",
    "    \"cv\": ShuffleSplit(n_splits=10, test_size=0.3, random_state=42),  #provide the cross-validation strategy\n",
    "    \"scoring\": \"explained_variance\",            # Scoring metric\n",
    "    \"n_jobs\": -1,                               # Use all CPU cores\n",
    "    \"line_kw\": {\"marker\": \"o\", \"alpha\": 0.8},   # Plot appearance\n",
    "    \"std_display_style\": \"fill_between\",        # provide the style for the standard deviation\n",
    "}\n",
    "\n",
    "# Create and plot learning curve\n",
    "fig, ax = plt.subplots(figsize=(10, 4))\n",
    "LearningCurveDisplay.from_estimator(\n",
    "    candidate_model,\n",
    "    **learning_curve_params,\n",
    "    ax=ax\n",
    ")\n",
    "\n",
    "# Customize the legend labels\n",
    "handles, labels = ax.get_legend_handles_labels()                         # Get the current legend handles and labels\n",
    "ax.legend(handles, [\"Training Metrics\", \"Testing Metrics\"], loc=\"best\")  # Update labels\n",
    "\n",
    "\n",
    "ax.set_title(\"Learning curve for the candidate model\")\n",
    "ax.set_xlabel(\"Number of samples\")\n",
    "ax.set_ylabel(\"Explained variance\")\n",
    "ax.set_ylim(0.8, 1.02)  # adjust ylimit\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- **Train Score**: as observed by the blue trend line, the train model performs well with a small number of samples. The decline in training score when the model is exposed to more data suggests that it's less able to fit the training set perfectly, likely due to increasing variance in the data.\n",
    "- **Cross Validation Score**: as can be observed by the orange line, it starts low but shows gradual increase when sample size increases, which is a good sign of generalisability. However, the gap between the two lines indicates that the training model is likely to overfit the train data, one could remove outliers in train set and observe the model performance change \n",
    "- **Converging Trend**: this indicates adding more data could improve model generalisability,and that the model is learning effectively. The plateau trend also indicated there is no meaningful improvement of model performance when sample size increases. One could add more data to training set, and continuously track the CV score. Further improvement may require feature engineering or a more expressive model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [8] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [8] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [8] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [8] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [2] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [2, 3] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [8] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [8] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [3, 8] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [3] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [8] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [8] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [8] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [1] in column 0 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [8] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [3] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [2] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [3] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [1] in column 0 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [9] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [1] in column 0 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [9] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n",
      "/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py:824: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "Traceback (most recent call last):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 813, in _score\n",
      "    scores = scorer(estimator, X_test, y_test)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/metrics/_scorer.py\", line 527, in __call__\n",
      "    return estimator.score(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 749, in score\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 816, in transform\n",
      "    Xs = self._fit_transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py\", line 670, in _fit_transform\n",
      "    return Parallel(n_jobs=self.n_jobs)(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 65, in __call__\n",
      "    return super().__call__(iterable_with_config)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 1088, in __call__\n",
      "    while self.dispatch_one_batch(iterator):\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 901, in dispatch_one_batch\n",
      "    self._dispatch(tasks)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 819, in _dispatch\n",
      "    job = self._backend.apply_async(batch, callback=cb)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 208, in apply_async\n",
      "    result = ImmediateResult(func)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/_parallel_backends.py\", line 597, in __init__\n",
      "    self.results = batch()\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in __call__\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/joblib/parallel.py\", line 288, in <listcomp>\n",
      "    return [func(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/parallel.py\", line 127, in __call__\n",
      "    return self.function(*args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 933, in _transform_one\n",
      "    res = transformer.transform(X)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/pipeline.py\", line 689, in transform\n",
      "    Xt = transform.transform(Xt)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/utils/_set_output.py\", line 140, in wrapped\n",
      "    data_to_wrap = f(self, X, *args, **kwargs)\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 1016, in transform\n",
      "    X_int, X_mask = self._transform(\n",
      "  File \"/opt/anaconda3/envs/Anacoda/lib/python3.10/site-packages/sklearn/preprocessing/_encoders.py\", line 199, in _transform\n",
      "    raise ValueError(msg)\n",
      "ValueError: Found unknown categories [8] in column 1 during transform\n",
      "\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "warnings.filterwarnings(\"ignore\")\n",
    "# Complexity analysis\n",
    "complexity_analysis_params = {\n",
    "    \"X\": df_train[FEATURES_CANDIDATE],\n",
    "    \"y\": df_train[\"log(Price)\"],\n",
    "    \"train_sizes\": np.linspace(0.1, 1.0, 10),\n",
    "    \"cv\": ShuffleSplit(n_splits=50, test_size=0.3, random_state=42),\n",
    "    \"n_jobs\": -1,\n",
    "    \"return_times\": True,\n",
    "}\n",
    "\n",
    "train_sizes, _, test_scores, fit_times, score_times = learning_curve(\n",
    "    candidate_model, **complexity_analysis_params\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create subplots for scalability analysis\n",
    "fig, axes = plt.subplots(nrows=2, ncols=1, figsize=(8, 6), sharex=True)\n",
    "\n",
    "# Plot scalability for fit time\n",
    "axes[0].plot(train_sizes, fit_times.mean(axis=1), \"o-\", label=\"Mean Fit Time\")\n",
    "axes[0].fill_between(\n",
    "    train_sizes,\n",
    "    fit_times.mean(axis=1) - fit_times.std(axis=1),\n",
    "    fit_times.mean(axis=1) + fit_times.std(axis=1),\n",
    "    alpha=0.3,\n",
    "    label=\"Fit Time ± Std Dev\"\n",
    ")\n",
    "axes[0].set_ylabel(\"Fit time (s)\")\n",
    "axes[0].set_title(\"Scalability of the Candidate Model (Fit Time)\")\n",
    "axes[0].legend()\n",
    "\n",
    "# Plot scalability for score time\n",
    "axes[1].plot(train_sizes, score_times.mean(axis=1), \"o-\", label=\"Mean Score Time\")\n",
    "axes[1].fill_between(\n",
    "    train_sizes,\n",
    "    score_times.mean(axis=1) - score_times.std(axis=1),\n",
    "    score_times.mean(axis=1) + score_times.std(axis=1),\n",
    "    alpha=0.3,\n",
    "    label=\"Score Time ± Std Dev\"\n",
    ")\n",
    "axes[1].set_ylabel(\"Score time (s)\")\n",
    "axes[1].set_xlabel(\"Number of training samples\")\n",
    "axes[1].legend()\n",
    "\n",
    "# Adjust layout and show plot\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<!-- - **Fit Time Plot**: The fit time increases predictably as the number of training samples grows, indicating the model scales linearly with data size. The small variation across folds suggests stable and consistent training performance.\n",
    "- **Score Time Plot**: The score time remains very low and stable across all training sample sizes, showing the model is highly efficient in making predictions, with minimal variability across cross-validation folds. -->"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**•** Once the model performance is satisfactory, measure the generalization error using the *hold-out* / *test* set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "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>Price</th>\n",
       "      <th>Address</th>\n",
       "      <th>Zip</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>Lon</th>\n",
       "      <th>Lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>325000.0</td>\n",
       "      <td>Nijenburg 20, Amsterdam</td>\n",
       "      <td>1081 GG</td>\n",
       "      <td>55</td>\n",
       "      <td>2</td>\n",
       "      <td>4.865786</td>\n",
       "      <td>52.328459</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>300000.0</td>\n",
       "      <td>Bolestein 756, Amsterdam</td>\n",
       "      <td>1081 EP</td>\n",
       "      <td>44</td>\n",
       "      <td>2</td>\n",
       "      <td>4.859323</td>\n",
       "      <td>52.324179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1050000.0</td>\n",
       "      <td>Stadhouderskade 16 H, Amsterdam</td>\n",
       "      <td>1054 ES</td>\n",
       "      <td>151</td>\n",
       "      <td>4</td>\n",
       "      <td>4.880571</td>\n",
       "      <td>52.362038</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>450000.0</td>\n",
       "      <td>Pruimenstraat 18 B, Amsterdam</td>\n",
       "      <td>1033 KM</td>\n",
       "      <td>87</td>\n",
       "      <td>3</td>\n",
       "      <td>4.896536</td>\n",
       "      <td>52.410585</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>350000.0</td>\n",
       "      <td>Frissenstein 304, Amsterdam</td>\n",
       "      <td>1102 AP</td>\n",
       "      <td>98</td>\n",
       "      <td>4</td>\n",
       "      <td>4.952422</td>\n",
       "      <td>52.316855</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Price                          Address      Zip  Area  Room       Lon  \\\n",
       "0   325000.0          Nijenburg 20, Amsterdam  1081 GG    55     2  4.865786   \n",
       "1   300000.0         Bolestein 756, Amsterdam  1081 EP    44     2  4.859323   \n",
       "2  1050000.0  Stadhouderskade 16 H, Amsterdam  1054 ES   151     4  4.880571   \n",
       "3   450000.0    Pruimenstraat 18 B, Amsterdam  1033 KM    87     3  4.896536   \n",
       "4   350000.0      Frissenstein 304, Amsterdam  1102 AP    98     4  4.952422   \n",
       "\n",
       "         Lat  \n",
       "0  52.328459  \n",
       "1  52.324179  \n",
       "2  52.362038  \n",
       "3  52.410585  \n",
       "4  52.316855  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "• List of features for model evaluation: ['log(Area)', 'log(Room)', 'log(AR-ratio)', 'Manhattan_distance', 'Region', 'Location']\n",
      "• Nans in Target variable: 1\n"
     ]
    }
   ],
   "source": [
    "# Load the test data\n",
    "DATA_TEST_PATH = os.path.join(\"data\", \"data_test.csv\")\n",
    "df_test = pd.read_csv(DATA_TEST_PATH)\n",
    "\n",
    "display(df_test.head())\n",
    "# list of features for model evaluation\n",
    "print(f\"• List of features for model evaluation: {FEATURES_CANDIDATE}\")\n",
    "print(f\"• Nans in Target variable: {np.isnan(df_test[TARGET]).sum()}\")\n",
    "\n",
    "# Drop row with NaN in the target variable\n",
    "df_test = df_test.dropna(subset=[TARGET])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Count NaNs post processing: log(Area)             0\n",
      "log(Room)             0\n",
      "log(AR-ratio)         0\n",
      "Manhattan_distance    0\n",
      "Region                0\n",
      "Location              0\n",
      "dtype: int64\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>log(Area)</th>\n",
       "      <th>log(Room)</th>\n",
       "      <th>log(AR-ratio)</th>\n",
       "      <th>Manhattan_distance</th>\n",
       "      <th>Region</th>\n",
       "      <th>Location</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4.007333</td>\n",
       "      <td>0.693147</td>\n",
       "      <td>3.314186</td>\n",
       "      <td>0.071905</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3.784190</td>\n",
       "      <td>0.693147</td>\n",
       "      <td>3.091042</td>\n",
       "      <td>0.082648</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5.017280</td>\n",
       "      <td>1.386294</td>\n",
       "      <td>3.630985</td>\n",
       "      <td>0.023541</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.465908</td>\n",
       "      <td>1.098612</td>\n",
       "      <td>3.367296</td>\n",
       "      <td>0.040971</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.584967</td>\n",
       "      <td>1.386294</td>\n",
       "      <td>3.198673</td>\n",
       "      <td>0.115607</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   log(Area)  log(Room)  log(AR-ratio)  Manhattan_distance  Region  Location\n",
       "0   4.007333   0.693147       3.314186            0.071905       0         8\n",
       "1   3.784190   0.693147       3.091042            0.082648       0         8\n",
       "2   5.017280   1.386294       3.630985            0.023541       0         5\n",
       "3   4.465908   1.098612       3.367296            0.040971       0         3\n",
       "4   4.584967   1.386294       3.198673            0.115607       1         0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Count NaNs post processing\n"
     ]
    }
   ],
   "source": [
    "# Perform all the feature engineering steps\n",
    "\n",
    "y_test = df_test[\"Price\"]\n",
    "\n",
    "# Define Lat and Lon of Dam Square\n",
    "Lat_Centrum = 52.373095\n",
    "Lon_Centrum = 4.893055\n",
    "\n",
    "def manhattan_distance(lat1: float, lon1: float, lat2: float, lon2: float) -> float:\n",
    "    \"\"\"\n",
    "    Calculate the distances between the (Lat, Lon) coordinates\n",
    "    \"\"\"\n",
    "\n",
    "    return np.abs(lat1 - lat2) + np.abs(lon1 - lon2)\n",
    "\n",
    "X_test = (\n",
    "    df_test.assign(\n",
    "        **{\n",
    "            \"log(Area)\": lambda x: np.log(x[\"Area\"]),\n",
    "            \"log(Room)\": lambda x: np.log(x[\"Room\"]),\n",
    "            \"log(AR-ratio)\": lambda x: np.log(x[\"Area\"] / x[\"Room\"]),\n",
    "            \"Manhattan_distance\": lambda x: manhattan_distance(\n",
    "                x[\"Lat\"], x[\"Lon\"], Lat_Centrum, Lon_Centrum\n",
    "            ),\n",
    "            \"Region\": lambda df: df[\"Zip\"].str[1].astype(int),  # second digit\n",
    "            \"Location\": lambda df: df[\"Zip\"].str[2].astype(int),  # third digit\n",
    "        }\n",
    "    )\n",
    "    # select only the columns of interest\n",
    "    .filter(FEATURES_CANDIDATE)\n",
    "    # safety step. Make sure the order of the columns is the same as the one used for training\n",
    "    .loc[:, FEATURES_CANDIDATE]\n",
    "    # count NaNs\n",
    "    .pipe(chain_snap, fn=lambda x: x.isna().sum(), msg=\"Count NaNs post processing\")\n",
    ")\n",
    "\n",
    "display(X_test.head())\n",
    "print(\"Count NaNs post processing\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Nans in y_pred_test: 0\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the model on the test set\n",
    "y_pred_test = np.exp(candidate_model.predict(X_test))\n",
    "\n",
    "# count NaNs\n",
    "print(f\"Nans in y_pred_test: {np.isnan(y_pred_test).sum()}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Candidate Model performance in Test Set\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>Score</th>\n",
       "      <th>Description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>r-squared</th>\n",
       "      <td>0.822</td>\n",
       "      <td>Coefficient of Determination</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>explained variance</th>\n",
       "      <td>0.825</td>\n",
       "      <td>Explained Variance Score</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>280684.906</td>\n",
       "      <td>Root Mean Squared Error (RMSE)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>116912.534</td>\n",
       "      <td>Mean Absolute Error (MAE)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max error</th>\n",
       "      <td>2776951.084</td>\n",
       "      <td>The maximum residual error</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          Score                     Description\n",
       "r-squared                 0.822    Coefficient of Determination\n",
       "explained variance        0.825        Explained Variance Score\n",
       "RMSE                 280684.906  Root Mean Squared Error (RMSE)\n",
       "MAE                  116912.534       Mean Absolute Error (MAE)\n",
       "max error           2776951.084      The maximum residual error"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "• Number of samples in Test Set = 184\n"
     ]
    }
   ],
   "source": [
    "print(\"Candidate Model performance in Test Set\")\n",
    "display(regression_report(y_true=y_test, y_pred=y_pred_test))\n",
    "print(f\"• Number of samples in Test Set = {len(df_test)}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Based on the above learnings on minimising the prediction errors with removal of outliers, one can expect with the same model handling normal Dutch houses (i.e., area <=128, and rooms between 2&4), the model would yield to a lower explained variance (similar to what we observe in validation set)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the criteria for filtering outliers\n",
    "def filter_outliers(data, area_col='Area', room_col='Room'):\n",
    "    \"\"\"\n",
    "    Filters the data based on the given criteria for Area and Room.\n",
    "    Args:\n",
    "        data (pd.DataFrame): Input dataset to filter.\n",
    "        area_col (str): Column name for Area.\n",
    "        room_col (str): Column name for Room.\n",
    "    Returns:\n",
    "        pd.DataFrame: Filtered dataset without outliers.\n",
    "        pd.Series: Boolean mask of non-outlier data.\n",
    "    \"\"\"\n",
    "    mask = (data[area_col] <= 128) & (data[room_col] >= 2) & (data[room_col] <= 4)\n",
    "    return data[mask], mask\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Count NaNs post processing: log(Area)             0\n",
      "log(Room)             0\n",
      "log(AR-ratio)         0\n",
      "Manhattan_distance    0\n",
      "Region                0\n",
      "Location              0\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# Filter test data\n",
    "filtered_test, test_mask = filter_outliers(df_test)\n",
    "\n",
    "# feature engineer for the filtered test data\n",
    "filtered_test = (\n",
    "    filtered_test.assign(\n",
    "        **{\n",
    "            \"log(Area)\": lambda x: np.log(x[\"Area\"]),\n",
    "            \"log(Room)\": lambda x: np.log(x[\"Room\"]),\n",
    "            \"log(AR-ratio)\": lambda x: np.log(x[\"Area\"] / x[\"Room\"]),\n",
    "            \"Manhattan_distance\": lambda x: manhattan_distance(\n",
    "                x[\"Lat\"], x[\"Lon\"], Lat_Centrum, Lon_Centrum\n",
    "            ),\n",
    "            \"Region\": lambda df: df[\"Zip\"].str[1].astype(int),  # second digit\n",
    "            \"Location\": lambda df: df[\"Zip\"].str[2].astype(int),  # third digit\n",
    "        }\n",
    "    )\n",
    "    # select only the columns of interest\n",
    "    .filter(FEATURES_CANDIDATE)\n",
    "    # safety step. Make sure the order of the columns is the same as the one used for training\n",
    "    .loc[:, FEATURES_CANDIDATE]\n",
    "    # count NaNs\n",
    "    .pipe(chain_snap, fn=lambda x: x.isna().sum(), msg=\"Count NaNs post processing\")\n",
    ")\n",
    "\n",
    "# Evaluate the model on the filtered test set\n",
    "y_pred_test_filtered = np.exp(candidate_model.predict(filtered_test))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Candidate Model performance in Test Set (filtered)\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>Score</th>\n",
       "      <th>Description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>r-squared</th>\n",
       "      <td>0.752</td>\n",
       "      <td>Coefficient of Determination</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>explained variance</th>\n",
       "      <td>0.752</td>\n",
       "      <td>Explained Variance Score</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>98593.267</td>\n",
       "      <td>Root Mean Squared Error (RMSE)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>66365.568</td>\n",
       "      <td>Mean Absolute Error (MAE)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max error</th>\n",
       "      <td>509406.169</td>\n",
       "      <td>The maximum residual error</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         Score                     Description\n",
       "r-squared                0.752    Coefficient of Determination\n",
       "explained variance       0.752        Explained Variance Score\n",
       "RMSE                 98593.267  Root Mean Squared Error (RMSE)\n",
       "MAE                  66365.568       Mean Absolute Error (MAE)\n",
       "max error           509406.169      The maximum residual error"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "• Number of samples in Test Set(filtered) = 136\n"
     ]
    }
   ],
   "source": [
    "print(\"Candidate Model performance in Test Set (filtered)\")\n",
    "display(regression_report(y_true=y_test[test_mask], y_pred=y_pred_test_filtered))\n",
    "print(f\"• Number of samples in Test Set(filtered) = {len(filtered_test)}\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Anacoda",
   "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": "3.10.8"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}