7_Cross_validation_Timeseries.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "0dbd0d92-196c-4b87-8a50-51b44ff572d1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load libraries\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import normalize, StandardScaler\n",
    "from sklearn.model_selection import GridSearchCV, RandomizedSearchCV\n",
    "from sklearn.metrics import mean_squared_error, confusion_matrix, classification_report, roc_auc_score, accuracy_score, roc_curve, auc, recall_score, precision_score, f1_score, precision_recall_curve\n",
    "from sklearn.model_selection import StratifiedKFold,train_test_split,KFold\n",
    "from sklearn.model_selection import cross_val_score, KFold\n",
    "from sklearn.model_selection import cross_validate\n",
    "from sklearn.preprocessing import label_binarize\n",
    "from sklearn import preprocessing\n",
    "\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.experimental import enable_iterative_imputer\n",
    "from sklearn.impute import IterativeImputer\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.linear_model import Lasso, LinearRegression, Ridge\n",
    "\n",
    "from sklearn.linear_model import LogisticRegression \n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from xgboost import XGBClassifier\n",
    "from sklearn.ensemble import BaggingClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier, VotingClassifier\n",
    "from sklearn.ensemble import AdaBoostClassifier\n",
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "from sklearn.multiclass import OneVsRestClassifier\n",
    "\n",
    "from sklearn.multioutput import MultiOutputClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "\n",
    "from catboost import Pool,CatBoostClassifier\n",
    "import warnings\n",
    "\n",
    "import shap\n",
    "from pingouin import partial_corr\n",
    "\n",
    "from sklearn.feature_selection import SelectKBest, chi2, f_classif\n",
    "\n",
    "from math import pi\n",
    "\n",
    "from sklearn.linear_model import LinearRegression, Ridge\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "import operator\n",
    "from functools import reduce"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "ee767023-975b-47b5-8eb2-572f760b2036",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load the data\n",
    "\n",
    "df = pd.read_csv(\"C:/Users/Meyke/WiDS/data/data_1101.csv\")\n",
    "df = df.set_index('id')\n",
    "\n",
    "df_data = pd.read_csv(\"C:/Users/Meyke/WiDS/data/train.csv\")\n",
    "df_test = pd.read_csv(\"C:/Users/Meyke/WiDS/data/test.csv\")\n",
    "\n",
    "df_data = df_data.set_index('id')\n",
    "df_test = df_test.set_index('id')\n",
    "\n",
    "index_test = df_test.index.to_list()\n",
    "index_train = df_data.index.to_list()\n",
    "\n",
    "df_train = df.iloc[index_train]\n",
    "df_test = df.iloc[index_test]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e4db6c5a-9cc7-4de3-bdb5-257701f68aee",
   "metadata": {},
   "source": [
    "### Prep data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "4979af90-c5d8-4236-97bc-6e31ddb790bf",
   "metadata": {},
   "outputs": [],
   "source": [
    "cat_cols = np.where(df.dtypes == object)[0]\n",
    "cat_cols = [df.columns[x] for x in cat_cols]\n",
    "\n",
    "numeric_cols = np.where(df.dtypes != object)[0]\n",
    "numeric_cols = [df.columns[x] for x in numeric_cols]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "845727e2-e81b-40ec-8571-413444583b48",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train = df_train.fillna(-999)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "311df647-a8ae-4457-b5c7-54209802642b",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train = df_train[numeric_cols]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "bdea3185-fbd1-4b7b-867b-ec7d5ec7ac69",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train_whatever = df_train"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b9df5c31-79dd-4ba9-ad32-033981237795",
   "metadata": {},
   "source": [
    "### Create a list of lists on year indexes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "3f87a0a1-c53a-47d8-afdc-170b4a1209e8",
   "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>building_id</th>\n",
       "      <th>building_id_singleton</th>\n",
       "      <th>year_factor</th>\n",
       "      <th>floor_area</th>\n",
       "      <th>year_built</th>\n",
       "      <th>energy_star_rating</th>\n",
       "      <th>elevation</th>\n",
       "      <th>january_min_temp</th>\n",
       "      <th>january_avg_temp</th>\n",
       "      <th>january_max_temp</th>\n",
       "      <th>...</th>\n",
       "      <th>days_below_0F</th>\n",
       "      <th>days_above_80F</th>\n",
       "      <th>days_above_90F</th>\n",
       "      <th>days_above_100F</th>\n",
       "      <th>days_above_110F</th>\n",
       "      <th>direction_max_wind_speed</th>\n",
       "      <th>direction_peak_wind_speed</th>\n",
       "      <th>max_wind_speed</th>\n",
       "      <th>days_with_fog</th>\n",
       "      <th>site_eui</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>313.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>61242.0</td>\n",
       "      <td>1942.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>2.4</td>\n",
       "      <td>36</td>\n",
       "      <td>50.5</td>\n",
       "      <td>68</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>14</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-999.0</td>\n",
       "      <td>248.682615</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>201.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>274000.0</td>\n",
       "      <td>1955.0</td>\n",
       "      <td>45.0</td>\n",
       "      <td>1.8</td>\n",
       "      <td>36</td>\n",
       "      <td>50.5</td>\n",
       "      <td>68</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>14</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-999.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>26.500150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>463.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>280025.0</td>\n",
       "      <td>1951.0</td>\n",
       "      <td>97.0</td>\n",
       "      <td>1.8</td>\n",
       "      <td>36</td>\n",
       "      <td>50.5</td>\n",
       "      <td>68</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>14</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-999.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>24.693619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>303.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>55325.0</td>\n",
       "      <td>1980.0</td>\n",
       "      <td>46.0</td>\n",
       "      <td>1.8</td>\n",
       "      <td>36</td>\n",
       "      <td>50.5</td>\n",
       "      <td>68</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>14</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-999.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>48.406926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>445.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>66000.0</td>\n",
       "      <td>1985.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>2.4</td>\n",
       "      <td>36</td>\n",
       "      <td>50.5</td>\n",
       "      <td>68</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>14</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-999.0</td>\n",
       "      <td>3.899395</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 62 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    building_id  building_id_singleton  year_factor  floor_area  year_built  \\\n",
       "id                                                                            \n",
       "0         313.0                      0            1     61242.0      1942.0   \n",
       "1         201.0                      0            1    274000.0      1955.0   \n",
       "2         463.0                      0            1    280025.0      1951.0   \n",
       "3         303.0                      0            1     55325.0      1980.0   \n",
       "4         445.0                      0            1     66000.0      1985.0   \n",
       "\n",
       "    energy_star_rating  elevation  january_min_temp  january_avg_temp  \\\n",
       "id                                                                      \n",
       "0                 11.0        2.4                36              50.5   \n",
       "1                 45.0        1.8                36              50.5   \n",
       "2                 97.0        1.8                36              50.5   \n",
       "3                 46.0        1.8                36              50.5   \n",
       "4                100.0        2.4                36              50.5   \n",
       "\n",
       "    january_max_temp  ...  days_below_0F  days_above_80F  days_above_90F  \\\n",
       "id                    ...                                                  \n",
       "0                 68  ...              0              14               0   \n",
       "1                 68  ...              0              14               0   \n",
       "2                 68  ...              0              14               0   \n",
       "3                 68  ...              0              14               0   \n",
       "4                 68  ...              0              14               0   \n",
       "\n",
       "    days_above_100F  days_above_110F  direction_max_wind_speed  \\\n",
       "id                                                               \n",
       "0                 0                0                       1.0   \n",
       "1                 0                0                       1.0   \n",
       "2                 0                0                       1.0   \n",
       "3                 0                0                       1.0   \n",
       "4                 0                0                       1.0   \n",
       "\n",
       "    direction_peak_wind_speed  max_wind_speed  days_with_fog    site_eui  \n",
       "id                                                                        \n",
       "0                         1.0             1.0         -999.0  248.682615  \n",
       "1                      -999.0             1.0           12.0   26.500150  \n",
       "2                      -999.0             1.0           12.0   24.693619  \n",
       "3                      -999.0             1.0           12.0   48.406926  \n",
       "4                         1.0             1.0         -999.0    3.899395  \n",
       "\n",
       "[5 rows x 62 columns]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#create a list of year indexes\n",
    "#df_train_whatever = df_train_whatever.set_index('id')\n",
    "\n",
    "listOfIndexes = df_train_whatever.index.to_list()\n",
    "listOfYears = df_train_whatever['year_factor'].to_list()\n",
    "\n",
    "list_of_lists=[[],[],[],[],[],[]]\n",
    "for index, value in enumerate(listOfYears):\n",
    "    list_of_lists[value-1].append(listOfIndexes[index])\n",
    "\n",
    "df_train_whatever.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ee641492-5fb4-45c0-b11c-9420a4764614",
   "metadata": {},
   "source": [
    "## Rolling forecast"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "ba5b7a81-34b7-480e-9a68-f98c9287b1ca",
   "metadata": {},
   "outputs": [],
   "source": [
    "def rolling_forecast_origin(train, min_train_size, horizon):\n",
    "    for i in range(len(train) - min_train_size - horizon + 1):\n",
    "        split_train = train[:min_train_size+i]\n",
    "        split_val = train[min_train_size+i:min_train_size+i+horizon]\n",
    "        yield split_train, split_val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "4bded5f5-f09c-4de2-a149-1310374433bb",
   "metadata": {},
   "outputs": [],
   "source": [
    "#cv_rolling = rolling_forecast_origin(train, min_train_size=4, horizon=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "85866f51-59d3-4a9c-ba7a-b00c98bacb02",
   "metadata": {},
   "source": [
    "## Sliding window"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "07b88d6f-3926-450d-8f76-2ab028df58c5",
   "metadata": {},
   "outputs": [],
   "source": [
    "def sliding_window(train, window_size, horizon):\n",
    "    for i in range(len(train) -  window_size - horizon +1):\n",
    "        split_train = train[i : window_size + i]\n",
    "        split_val = train[i + window_size : window_size + i + horizon]\n",
    "        yield split_train, split_val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "c23adf02-ffcd-411f-9338-7c15a5b3c8b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "cv_sliding =  sliding_window(list_of_lists, window_size = 3, horizon = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "03a10148-6336-4c28-90c7-f95688f4376a",
   "metadata": {},
   "outputs": [],
   "source": [
    "import random"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ae5bd201-662f-4ff5-ab4a-f2f3ba9ada3a",
   "metadata": {},
   "source": [
    "## Example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "c57b482a-c7a1-4ea4-9a87-eb0634a8b141",
   "metadata": {},
   "outputs": [],
   "source": [
    "#create cross validation function\n",
    "def cross_validation_score(model, train, cv, metric):\n",
    "    cv_scores = []\n",
    "    pred_df = pd.DataFrame({'ids': [],\n",
    "                    'true': [],\n",
    "                    'pred': []})\n",
    "    for cv_train, cv_test in cv:\n",
    "        \n",
    "        # flatten the list of lists \n",
    "        cv_train = reduce(operator.concat, cv_train)\n",
    "        cv_test = reduce(operator.concat, cv_test)\n",
    "        \n",
    "        # calc the test size\n",
    "        len_cv_test = round(len(cv_train)*0.25)\n",
    "        \n",
    "        for x in range(5): \n",
    "            cv_test_split = []\n",
    "            scores = []\n",
    "            for size in range(len_cv_test):\n",
    "                cv_test_split.append(random.randrange(len(cv_test)))\n",
    "            \n",
    "        \n",
    "        \n",
    "            # define a list of y_trues\n",
    "            y_true = df_train_whatever.iloc[cv_test_split]['site_eui'].values.tolist()\n",
    "\n",
    "            # fit the model\n",
    "            model.fit(df_train_whatever.iloc[cv_train,:].drop(columns=[\"site_eui\"]), df_train_whatever.iloc[cv_train,:]['site_eui'])\n",
    "\n",
    "            # create predictions\n",
    "            preds = model.predict(df_train_whatever.iloc[cv_test_split].drop(columns=[\"site_eui\"]))   #zou mooi zijn om deze niet hard te encoden, zodat we hem ook kunnen vervangen door log_site_eui\n",
    "            preds = preds.tolist()\n",
    "        \n",
    "            # calculate score and append to cv_scores\n",
    "            score = metric(y_true, preds, squared=False) \n",
    "            scores.append(score)\n",
    "        \n",
    "            # append df with predictions and true labels for this split\n",
    "            df_temp = pd.DataFrame({'ids': cv_test_split,\n",
    "                        'true': y_true,\n",
    "                        'pred': preds})\n",
    "            pred_df = pred_df.append(df_temp, ignore_index=True)\n",
    "        \n",
    "        cv_scores.append(sum(scores) / len(scores))\n",
    "        \n",
    "    return cv_scores, pred_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "e66ed70b-a2f2-4df2-a30a-098d5995d8ca",
   "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>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>7.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>61.248236</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.645349</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>58.894734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>60.449189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>61.301283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>61.920673</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>63.801913</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               0\n",
       "count   7.000000\n",
       "mean   61.248236\n",
       "std     1.645349\n",
       "min    58.894734\n",
       "25%    60.449189\n",
       "50%    61.301283\n",
       "75%    61.920673\n",
       "max    63.801913"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "alpha = [0.001, 0.01, 0.1, 1, 10, 100, 1000]\n",
    "\n",
    "cv_scores = []\n",
    "for estimator in alpha:\n",
    "    cv_sliding =  sliding_window(list_of_lists, window_size = 5, horizon = 1)\n",
    "    cv_scores_2, df = cross_validation_score(model=Ridge(alpha=estimator),\n",
    "                                    train=list_of_lists,\n",
    "                                    cv=cv_sliding,\n",
    "                                    metric=mean_squared_error)\n",
    "    cv_scores.append(cv_scores_2)\n",
    "\n",
    "pd.DataFrame(cv_scores).describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "b740920b-bab9-4faa-bee4-c6d75b3958ee",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[58.89473423056794],\n",
       " [62.47303600944742],\n",
       " [59.61597908803297],\n",
       " [63.80191341843062],\n",
       " [61.28239976919333],\n",
       " [61.301282885713775],\n",
       " [61.36830977073755]]"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "743daff4-fd12-4ca3-9c31-068779eba44b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# transpose and de-transpose df\n",
    "transpose_df = pd.DataFrame(cv_scores).T.describe()\n",
    "de_transpose_df = transpose_df.T\n",
    "\n",
    "# plot results\n",
    "fig, ax = plt.subplots(figsize=(6, 6))\n",
    "\n",
    "# setting the axis' labels\n",
    "ax.set_ylabel('Mean Absolute Error',fontsize=12)\n",
    "ax.set_xlabel('estimators',fontsize=12)\n",
    "ax.set_xticks(de_transpose_df.index.to_list())\n",
    "\n",
    "# plot a line for mean, max, min\n",
    "de_transpose_df[['mean','max','min']].plot(ax=ax);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e3edd42f-b279-4f28-8130-8c38ea238856",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "19f373ea-4025-471b-9a5f-c82bef00d9e0",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}