regression.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "1875f256-ec28-44c7-abb4-62003ad580f1",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from tqdm import tqdm\n",
    "from glob import glob\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from xgboost import XGBRegressor\n",
    "from sklearn.preprocessing import StandardScaler, LabelEncoder, PolynomialFeatures\n",
    "from sklearn.ensemble import RandomForestRegressor, GradientBoostingRegressor, AdaBoostRegressor\n",
    "from sklearn.metrics import mean_squared_error, mean_absolute_error, r2_score\n",
    "from sklearn.linear_model import LinearRegression, ElasticNet, Ridge, Lasso, RANSACRegressor\n",
    "from sklearn.model_selection import cross_val_score, cross_validate, KFold, StratifiedKFold, train_test_split, GridSearchCV\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.compose import TransformedTargetRegressor\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from yellowbrick.regressor import PredictionError, ResidualsPlot\n",
    "from feature_extraction import extract_features_from_images\n",
    "from sklearn.cross_decomposition import PLSRegression\n",
    "from yellowbrick.regressor import "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "id": "47ee1c44-3104-47c6-8ad5-90e632b7b483",
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "plt.style.use('default')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "f8c6fe0d-54ec-4e5c-8271-e63bd40823a9",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def handle_model_evaluate(true, predicted, title_column=\"\"):\n",
    "\n",
    "    mae = mean_absolute_error(true, predicted)\n",
    "    mse = mean_squared_error(true, predicted)\n",
    "    rmse = np.sqrt(mse)\n",
    "    r2_square = r2_score(true, predicted)\n",
    "    mape = np.mean(np.abs((true - predicted) / true)) * 100\n",
    "\n",
    "    return pd.DataFrame(\n",
    "        {\n",
    "            f\"{title_column}\": [mae, mse, rmse, mape, r2_square],\n",
    "        },\n",
    "        index=[\"MAE\", \"MSE\", \"RMSE\", \"MAPE\", \"R-squared\"],\n",
    "    )\n",
    "\n",
    "    \n",
    "def show_model_performance(model, X_train, y_train, X_test, y_test):\n",
    "    \n",
    "    train = handle_model_evaluate(y_train, model.predict(X_train), \"Treino\")\n",
    "    test  = handle_model_evaluate(y_test, model.predict(X_test), \"Teste\")\n",
    "    \n",
    "    display(pd.concat([train, test], axis=1))\n",
    "    \n",
    "    \n",
    "def run_all_algs(X_train, y_train, X_test, y_test):\n",
    "\n",
    "    models = {\n",
    "        \"GradientBoostingRegressor\": GradientBoostingRegressor(),\n",
    "        # \"AdaBoostRegressor\": AdaBoostRegressor(),\n",
    "        \"LinearRegression\": LinearRegression(),\n",
    "        \"LassoRegression\": Lasso(),\n",
    "        \"RidgeRegression\": Ridge(),\n",
    "        \"RandomForestRegressor\": RandomForestRegressor(),\n",
    "        # \"ElasticNet\": ElasticNet(),\n",
    "        # \"[RANSACRegressor]\": RANSACRegressor(),\n",
    "        # \"[PLSRegression]\": PLSRegression(n_components=2),\n",
    "        \"PCARegression\": make_pipeline(PCA(n_components=2), LinearRegression()),\n",
    "        \"PolyRegression\": make_pipeline(PolynomialFeatures(degree=3), Lasso()),\n",
    "        # \"XGBRegressor\": XGBRegressor()\n",
    "    }\n",
    "    \n",
    "    for name, model in models.items():\n",
    "\n",
    "        model.fit(X_train, y_train)\n",
    "        \n",
    "        print(\"\\n\",\"\\b\"+name)\n",
    "        \n",
    "        show_model_performance(model,\n",
    "                               X_train,\n",
    "                               y_train,\n",
    "                               X_test,\n",
    "                               y_test)\n",
    "\n",
    "        print(\"_\" * 20)\n",
    "    \n",
    "def print_evaluate(true, predicted, title_column=\"\"):  \n",
    "    \n",
    "    mae = mean_absolute_error(true, predicted)\n",
    "    mse = mean_squared_error(true, predicted)\n",
    "    rmse = np.sqrt(mse)\n",
    "    r2_square = r2_score(true, predicted)\n",
    "    mape = np.mean(np.abs((true - predicted) / true)) * 100\n",
    "    \n",
    "    print(pd.DataFrame({\n",
    "        f\"{title_column}\": [\n",
    "            mae, \n",
    "            mse, \n",
    "            rmse,\n",
    "            mape,\n",
    "            r2_square],\n",
    "    },\n",
    "        index=[\n",
    "            \"MAE\",\n",
    "            \"MSE\",\n",
    "            \"RMSE\",\n",
    "            \"MAPE\",\n",
    "            \"R-squared\"\n",
    "        ]))\n",
    "\n",
    "    \n",
    "def ApplyesKFoldV2(x, y, k=5):\n",
    "    metrics = [\n",
    "        \"r2\",\n",
    "        \"neg_mean_absolute_error\",\n",
    "        \"neg_mean_squared_error\",\n",
    "        \"neg_root_mean_squared_error\",\n",
    "        \"neg_mean_absolute_percentage_error\"    \n",
    "    ]\n",
    "    \n",
    "    kfold  = KFold(n_splits=k, shuffle=True, random_state=0)\n",
    "\n",
    "    models = {\n",
    "        \"GradientBoostingRegressor\": GradientBoostingRegressor(),\n",
    "        # \"AdaBoostRegressor\": AdaBoostRegressor(),\n",
    "        \"LinearRegression\": LinearRegression(),\n",
    "        \"LassoRegression\": Lasso(),\n",
    "        \"RidgeRegression\": Ridge(),\n",
    "        \"RandomForestRegressor\": RandomForestRegressor(),\n",
    "        # \"ElasticNet\": ElasticNet(),\n",
    "        # \"RANSACRegressor\": RANSACRegressor(),\n",
    "        # \"PLSRegression\": PLSRegression(n_components=2),\n",
    "        \"PCARegression\": make_pipeline(PCA(n_components=2), LinearRegression()),\n",
    "        \"PolyRegression\": make_pipeline(PolynomialFeatures(degree=2), Lasso()),\n",
    "    }\n",
    "\n",
    "    results = {}\n",
    "    \n",
    "    for name, model in models.items():\n",
    "        results[name] = cross_validate(model, x, y, cv=kfold, scoring=metrics)\n",
    "        \n",
    "    for name, cv_results in results.items():\n",
    "        print(f\"[{name}]\")\n",
    "        display(pd.DataFrame({\n",
    "            f\"Média\": [\n",
    "                cv_results['test_r2'].mean(), \n",
    "                -cv_results['test_neg_mean_absolute_error'].mean(),\n",
    "                -cv_results['test_neg_mean_squared_error'].mean(),\n",
    "                -cv_results['test_neg_root_mean_squared_error'].mean(),\n",
    "                -cv_results['test_neg_mean_absolute_percentage_error'].mean()\n",
    "            ],\n",
    "            \"Desvio Padrão\": [\n",
    "                cv_results['test_r2'].std(),\n",
    "                cv_results['test_neg_mean_absolute_error'].std(),\n",
    "                cv_results['test_neg_mean_squared_error'].std(),\n",
    "                cv_results['test_neg_root_mean_squared_error'].std(),\n",
    "                cv_results['test_neg_mean_absolute_percentage_error'].std()\n",
    "            ]\n",
    "        },\n",
    "            index=[\"R-squared\", \"MAE\", \"MSE\", \"RMSE\", \"MAPE\"],\n",
    "        ))\n",
    "\n",
    "\n",
    "def apply_pca_reduce(data, n=None):\n",
    "    \n",
    "    scaler = StandardScaler()\n",
    "    X = scaler.fit_transform(data)\n",
    "    \n",
    "    pca = PCA(n_components=data.shape[-1]).fit(X)\n",
    "    \n",
    "    cumsum = np.cumsum(pca.explained_variance_ratio_)\n",
    "    dim = np.argmax(cumsum >= 0.90) + 1\n",
    "    \n",
    "    print('Para preservar 90% da variance é necessário ' + str(dim))\n",
    "        \n",
    "    pca_refit = PCA(n_components=n if n else dim).fit(X)\n",
    "    print('Ratio: ', pca_refit.explained_variance_ratio_)\n",
    "    \n",
    "    return pca_refit.transform(X)\n",
    "\n",
    "def corr_view(df, w=8, h=6):\n",
    "    corrmat = df.corr()\n",
    "\n",
    "    _, ax = plt.subplots(figsize=(w, h))\n",
    "\n",
    "    sns.heatmap(corrmat, linewidths=5, annot=True);\n",
    "    \n",
    "    plt.show()\n",
    "    \n",
    "    return corrmat"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "de86c219-8b67-4c20-b13a-b9612fad42d0",
   "metadata": {},
   "source": [
    "## Extraindo Features das imagens mascaras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "6d86390b-aa3f-4a6b-8e29-145fcd64be18",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_dataset(base_path):\n",
    "    \n",
    "    *_, base_path_last = base_path.split(\"/\")\n",
    "    features_file_name = f\"{base_path_last}_features.csv\"\n",
    "    \n",
    "    if not os.path.isfile(features_file_name):\n",
    "        extract_features_from_images(\n",
    "            images_arr=[],\n",
    "            images_path=f\"{base_path}/masks\",\n",
    "            output_file_name=features_file_name,\n",
    "            resize_shape=(224, 224))\n",
    "    \n",
    "    weights = pd.read_csv(f\"{base_path}/annotations.csv\")\n",
    "    features = pd.read_csv(features_file_name)\n",
    "\n",
    "    weights.index = weights.img\n",
    "    features.index = features.label\n",
    "\n",
    "    weights = weights.drop('img', axis=1)\n",
    "    features = features.drop('label', axis=1)\n",
    "    \n",
    "    data = pd.concat([weights, features],\n",
    "                     axis=1,\n",
    "                     ignore_index=False, \n",
    "                     verify_integrity=True)\n",
    "\n",
    "    data = data.reset_index()\n",
    "    \n",
    "    return data\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "0615dc38-dad6-417b-bf5e-ccd1985de297",
   "metadata": {},
   "outputs": [],
   "source": [
    "data = load_dataset(\"data/coletas/combined\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "665d96d0-ab8a-44bb-8a4e-9ed87510d5b9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tambaqui    148\n",
       "tilápia      61\n",
       "curimatã     20\n",
       "panga         8\n",
       "traíra        5\n",
       "lambari       3\n",
       "Name: specie, dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.specie.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "fbb2d989-5fee-496d-93b9-e9f7e19955e7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'105.0 kg'"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f\"{data.weight.sum() // 1000} \" + \"kg\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "cd81eed5-ec2c-499d-b818-635cb7fb7a61",
   "metadata": {},
   "outputs": [],
   "source": [
    "# data.weight /= 1000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "479c6b3c-7e79-41c5-aa32-37ba684287e5",
   "metadata": {},
   "outputs": [],
   "source": [
    "enc = LabelEncoder()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d5ad1ce4-6d15-48a6-a510-9b19a1500f0a",
   "metadata": {},
   "outputs": [],
   "source": [
    "data.specie = enc.fit_transform(data.specie.values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "ef083dbd-bb9a-4b3b-b6f1-a31eda9948a9",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_unet_dataset(df):\n",
    "    \n",
    "    weights = df[[\"weight\", \"length\", \"specie\", \"width\"]]\n",
    "    weights.index = df[\"index\"]\n",
    "\n",
    "    features_file_name = \"features_from_unet_predictions.csv\"\n",
    "    features = pd.read_csv(features_file_name)\n",
    "    \n",
    "    features.index = features.label\n",
    "\n",
    "    features = features.drop('label', axis=1)\n",
    "    \n",
    "    data = pd.concat([weights, features],\n",
    "                     axis=1,\n",
    "                     ignore_index=False, \n",
    "                     verify_integrity=True)\n",
    "\n",
    "    data = data.reset_index()\n",
    "    \n",
    "    return data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3fba9d71-2b9d-40be-a4b2-cc41be12177f",
   "metadata": {},
   "source": [
    "### Ajustando Features Extraídas da U-net"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "235cd318-df00-45ae-ba17-e9957461604b",
   "metadata": {},
   "outputs": [],
   "source": [
    "_, X_test = train_test_split(data, test_size=0.3, random_state=0)\n",
    "unet_testing_features = load_unet_dataset(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "9558120a-5fc4-4ab6-b93c-255d4f5b887a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 100x100 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "corrmat = corr_view(data, w=1, h=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "3387c2f9-f71f-4f43-a949-431452cc6573",
   "metadata": {},
   "outputs": [],
   "source": [
    "features_names = list(dict(corrmat[corrmat.weight > 0.85].weight).keys())\n",
    "features_names.append(\"specie\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "7261dbd7-a9bc-47df-b56d-591751b925bf",
   "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>weight</th>\n",
       "      <th>length</th>\n",
       "      <th>area</th>\n",
       "      <th>bbox_area</th>\n",
       "      <th>convex_area</th>\n",
       "      <th>equivalent_diameter</th>\n",
       "      <th>filled_area</th>\n",
       "      <th>major_axis_length</th>\n",
       "      <th>moments-0-0</th>\n",
       "      <th>moments-0-1</th>\n",
       "      <th>...</th>\n",
       "      <th>moments-2-1</th>\n",
       "      <th>moments-2-2</th>\n",
       "      <th>moments-2-3</th>\n",
       "      <th>moments_central-0-0</th>\n",
       "      <th>moments_central-0-2</th>\n",
       "      <th>moments_central-2-0</th>\n",
       "      <th>moments_central-2-2</th>\n",
       "      <th>perimeter</th>\n",
       "      <th>feret_diameter_max</th>\n",
       "      <th>specie</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>220.0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>5867</td>\n",
       "      <td>8908</td>\n",
       "      <td>6372</td>\n",
       "      <td>86.429719</td>\n",
       "      <td>5867</td>\n",
       "      <td>122.598160</td>\n",
       "      <td>5867.0</td>\n",
       "      <td>344885.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4.695134e+08</td>\n",
       "      <td>3.385144e+10</td>\n",
       "      <td>2.780223e+12</td>\n",
       "      <td>5867.0</td>\n",
       "      <td>5.510758e+06</td>\n",
       "      <td>1.470126e+06</td>\n",
       "      <td>7.711683e+08</td>\n",
       "      <td>340.048773</td>\n",
       "      <td>132.230859</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>545.0</td>\n",
       "      <td>26.0</td>\n",
       "      <td>10098</td>\n",
       "      <td>15272</td>\n",
       "      <td>10689</td>\n",
       "      <td>113.389474</td>\n",
       "      <td>10098</td>\n",
       "      <td>154.971729</td>\n",
       "      <td>10098.0</td>\n",
       "      <td>818043.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.222528e+09</td>\n",
       "      <td>2.201669e+11</td>\n",
       "      <td>2.427110e+13</td>\n",
       "      <td>10098.0</td>\n",
       "      <td>1.513501e+07</td>\n",
       "      <td>4.568750e+06</td>\n",
       "      <td>3.687127e+09</td>\n",
       "      <td>438.374675</td>\n",
       "      <td>166.364059</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1025.0</td>\n",
       "      <td>31.0</td>\n",
       "      <td>13583</td>\n",
       "      <td>20580</td>\n",
       "      <td>14173</td>\n",
       "      <td>131.508223</td>\n",
       "      <td>13583</td>\n",
       "      <td>180.579492</td>\n",
       "      <td>13583.0</td>\n",
       "      <td>1285297.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3.716828e+09</td>\n",
       "      <td>4.027404e+11</td>\n",
       "      <td>4.892637e+13</td>\n",
       "      <td>13583.0</td>\n",
       "      <td>2.755749e+07</td>\n",
       "      <td>8.335128e+06</td>\n",
       "      <td>9.660429e+09</td>\n",
       "      <td>500.114790</td>\n",
       "      <td>196.918765</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>740.0</td>\n",
       "      <td>37.0</td>\n",
       "      <td>8521</td>\n",
       "      <td>12864</td>\n",
       "      <td>9204</td>\n",
       "      <td>104.159849</td>\n",
       "      <td>8521</td>\n",
       "      <td>183.209467</td>\n",
       "      <td>8521.0</td>\n",
       "      <td>791315.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.111224e+09</td>\n",
       "      <td>1.270943e+11</td>\n",
       "      <td>1.647843e+13</td>\n",
       "      <td>8521.0</td>\n",
       "      <td>1.787565e+07</td>\n",
       "      <td>1.987899e+06</td>\n",
       "      <td>2.195729e+09</td>\n",
       "      <td>459.948268</td>\n",
       "      <td>192.210822</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>475.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>8184</td>\n",
       "      <td>11907</td>\n",
       "      <td>8614</td>\n",
       "      <td>102.079344</td>\n",
       "      <td>8184</td>\n",
       "      <td>138.020198</td>\n",
       "      <td>8184.0</td>\n",
       "      <td>587517.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.239091e+09</td>\n",
       "      <td>1.098955e+11</td>\n",
       "      <td>1.083366e+13</td>\n",
       "      <td>8184.0</td>\n",
       "      <td>9.715353e+06</td>\n",
       "      <td>3.038349e+06</td>\n",
       "      <td>2.008774e+09</td>\n",
       "      <td>380.776695</td>\n",
       "      <td>147.275253</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 27 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   weight  length   area  bbox_area  convex_area  equivalent_diameter  \\\n",
       "0   220.0    22.0   5867       8908         6372            86.429719   \n",
       "1   545.0    26.0  10098      15272        10689           113.389474   \n",
       "2  1025.0    31.0  13583      20580        14173           131.508223   \n",
       "3   740.0    37.0   8521      12864         9204           104.159849   \n",
       "4   475.0    24.0   8184      11907         8614           102.079344   \n",
       "\n",
       "   filled_area  major_axis_length  moments-0-0  moments-0-1  ...  \\\n",
       "0         5867         122.598160       5867.0     344885.0  ...   \n",
       "1        10098         154.971729      10098.0     818043.0  ...   \n",
       "2        13583         180.579492      13583.0    1285297.0  ...   \n",
       "3         8521         183.209467       8521.0     791315.0  ...   \n",
       "4         8184         138.020198       8184.0     587517.0  ...   \n",
       "\n",
       "    moments-2-1   moments-2-2   moments-2-3  moments_central-0-0  \\\n",
       "0  4.695134e+08  3.385144e+10  2.780223e+12               5867.0   \n",
       "1  2.222528e+09  2.201669e+11  2.427110e+13              10098.0   \n",
       "2  3.716828e+09  4.027404e+11  4.892637e+13              13583.0   \n",
       "3  1.111224e+09  1.270943e+11  1.647843e+13               8521.0   \n",
       "4  1.239091e+09  1.098955e+11  1.083366e+13               8184.0   \n",
       "\n",
       "   moments_central-0-2  moments_central-2-0  moments_central-2-2   perimeter  \\\n",
       "0         5.510758e+06         1.470126e+06         7.711683e+08  340.048773   \n",
       "1         1.513501e+07         4.568750e+06         3.687127e+09  438.374675   \n",
       "2         2.755749e+07         8.335128e+06         9.660429e+09  500.114790   \n",
       "3         1.787565e+07         1.987899e+06         2.195729e+09  459.948268   \n",
       "4         9.715353e+06         3.038349e+06         2.008774e+09  380.776695   \n",
       "\n",
       "   feret_diameter_max  specie  \n",
       "0          132.230859       3  \n",
       "1          166.364059       4  \n",
       "2          196.918765       0  \n",
       "3          192.210822       2  \n",
       "4          147.275253       4  \n",
       "\n",
       "[5 rows x 27 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[features_names].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "15fba32f-c50a-4276-94cd-7d6b7ed5d48d",
   "metadata": {},
   "outputs": [],
   "source": [
    "unet_testing_features = unet_testing_features[features_names]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "23fcd4f0-002c-4252-872a-63b8ac877240",
   "metadata": {},
   "outputs": [],
   "source": [
    "y = data.weight\n",
    "X = data[features_names].drop(labels=['weight'], axis=1)\n",
    "l = data['index']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "3e936ab1-9427-4f6c-a238-b2c2d6e8567e",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test, index_train, index_test = train_test_split(X.values, \n",
    "                                                                             y.values,\n",
    "                                                                             l.values.tolist(),\n",
    "                                                                             test_size=0.3, \n",
    "                                                                             random_state=0)\n",
    "\n",
    "Unet_y_test = unet_testing_features.weight\n",
    "unet_testing_features = unet_testing_features[features_names].drop(labels=['weight'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "0fdedcdb-e9af-4bea-9d8c-2347451c35a8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(171, 74)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(X_train), len(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "2f268823-7f31-4d57-ae5f-3288fe4fcf53",
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "\n",
    "X_train_scaled, X_test_scaled = scaler.fit_transform(X_train), scaler.transform(X_test)\n",
    "Unet_X_test_scaled = scaler.transform(unet_testing_features.values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3eb95455-e47d-4644-90ec-a3d1f3167a18",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "5e9f73ef-e20b-46f4-b80c-c64de9b49456",
   "metadata": {},
   "source": [
    "# Features Normalizadas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "67b7be79-497f-4e91-bf54-745c388a9791",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "GradientBoostingRegressor\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>5.753149</td>\n",
       "      <td>34.080842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>53.400895</td>\n",
       "      <td>6913.339257</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>7.307592</td>\n",
       "      <td>83.146493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>6.286146</td>\n",
       "      <td>7.738139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.999181</td>\n",
       "      <td>0.881033</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Treino        Teste\n",
       "MAE         5.753149    34.080842\n",
       "MSE        53.400895  6913.339257\n",
       "RMSE        7.307592    83.146493\n",
       "MAPE        6.286146     7.738139\n",
       "R-squared   0.999181     0.881033"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "LinearRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>23.975293</td>\n",
       "      <td>41.857285</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>962.880039</td>\n",
       "      <td>8170.421887</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>31.030308</td>\n",
       "      <td>90.390386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>37.311182</td>\n",
       "      <td>13.513414</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.985241</td>\n",
       "      <td>0.859400</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Treino        Teste\n",
       "MAE         23.975293    41.857285\n",
       "MSE        962.880039  8170.421887\n",
       "RMSE        31.030308    90.390386\n",
       "MAPE        37.311182    13.513414\n",
       "R-squared    0.985241     0.859400"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "LassoRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>31.956953</td>\n",
       "      <td>46.104884</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>1804.489268</td>\n",
       "      <td>9151.278810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>42.479280</td>\n",
       "      <td>95.662317</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>21.587117</td>\n",
       "      <td>14.614081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.972340</td>\n",
       "      <td>0.842521</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Treino        Teste\n",
       "MAE          31.956953    46.104884\n",
       "MSE        1804.489268  9151.278810\n",
       "RMSE         42.479280    95.662317\n",
       "MAPE         21.587117    14.614081\n",
       "R-squared     0.972340     0.842521"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "RidgeRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>28.597508</td>\n",
       "      <td>42.378959</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>1448.371179</td>\n",
       "      <td>8878.617778</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>38.057472</td>\n",
       "      <td>94.226418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>15.823772</td>\n",
       "      <td>13.928071</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.977799</td>\n",
       "      <td>0.847213</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Treino        Teste\n",
       "MAE          28.597508    42.378959\n",
       "MSE        1448.371179  8878.617778\n",
       "RMSE         38.057472    94.226418\n",
       "MAPE         15.823772    13.928071\n",
       "R-squared     0.977799     0.847213"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "RandomForestRegressor\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>11.722678</td>\n",
       "      <td>34.963216</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>444.830216</td>\n",
       "      <td>6880.521693</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>21.090998</td>\n",
       "      <td>82.948910</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>41.852797</td>\n",
       "      <td>7.590665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.993182</td>\n",
       "      <td>0.881597</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Treino        Teste\n",
       "MAE         11.722678    34.963216\n",
       "MSE        444.830216  6880.521693\n",
       "RMSE        21.090998    82.948910\n",
       "MAPE        41.852797     7.590665\n",
       "R-squared    0.993182     0.881597"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "PCARegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>37.241035</td>\n",
       "      <td>50.885100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>2470.607808</td>\n",
       "      <td>9959.732323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>49.705209</td>\n",
       "      <td>99.798459</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>92.887716</td>\n",
       "      <td>15.452599</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.962130</td>\n",
       "      <td>0.828609</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Treino        Teste\n",
       "MAE          37.241035    50.885100\n",
       "MSE        2470.607808  9959.732323\n",
       "RMSE         49.705209    99.798459\n",
       "MAPE         92.887716    15.452599\n",
       "R-squared     0.962130     0.828609"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "PolyRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>19.584782</td>\n",
       "      <td>44.523812</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>622.533633</td>\n",
       "      <td>11046.601860</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>24.950624</td>\n",
       "      <td>105.102816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>28.739563</td>\n",
       "      <td>18.416496</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.990458</td>\n",
       "      <td>0.809906</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Treino         Teste\n",
       "MAE         19.584782     44.523812\n",
       "MSE        622.533633  11046.601860\n",
       "RMSE        24.950624    105.102816\n",
       "MAPE        28.739563     18.416496\n",
       "R-squared    0.990458      0.809906"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n"
     ]
    }
   ],
   "source": [
    "run_all_algs(X_train_scaled, \n",
    "             y_train,\n",
    "             X_test_scaled,\n",
    "             y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e50c42c9-ed71-4ac9-85a0-63daaa3ce9a0",
   "metadata": {},
   "source": [
    "# Features Não Normalizadas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "5a943f77-b834-4881-9326-e1e50b4cbcdf",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "GradientBoostingRegressor\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>5.753149</td>\n",
       "      <td>33.900804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>53.400895</td>\n",
       "      <td>6887.236139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>7.307592</td>\n",
       "      <td>82.989374</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>6.286146</td>\n",
       "      <td>7.606488</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.999181</td>\n",
       "      <td>0.881482</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Treino        Teste\n",
       "MAE         5.753149    33.900804\n",
       "MSE        53.400895  6887.236139\n",
       "RMSE        7.307592    82.989374\n",
       "MAPE        6.286146     7.606488\n",
       "R-squared   0.999181     0.881482"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "LinearRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>23.974582</td>\n",
       "      <td>41.854334</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>962.880129</td>\n",
       "      <td>8170.057059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>31.030310</td>\n",
       "      <td>90.388368</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>37.276645</td>\n",
       "      <td>13.512083</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.985241</td>\n",
       "      <td>0.859406</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Treino        Teste\n",
       "MAE         23.974582    41.854334\n",
       "MSE        962.880129  8170.057059\n",
       "RMSE        31.030310    90.388368\n",
       "MAPE        37.276645    13.512083\n",
       "R-squared    0.985241     0.859406"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "LassoRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>26.431473</td>\n",
       "      <td>39.656571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>1171.437241</td>\n",
       "      <td>8451.399781</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>34.226265</td>\n",
       "      <td>91.931495</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>27.699546</td>\n",
       "      <td>13.043333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.982044</td>\n",
       "      <td>0.854565</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Treino        Teste\n",
       "MAE          26.431473    39.656571\n",
       "MSE        1171.437241  8451.399781\n",
       "RMSE         34.226265    91.931495\n",
       "MAPE         27.699546    13.043333\n",
       "R-squared     0.982044     0.854565"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "RidgeRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>23.950637</td>\n",
       "      <td>41.683038</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>962.984248</td>\n",
       "      <td>8140.760631</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>31.031987</td>\n",
       "      <td>90.226164</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>36.381970</td>\n",
       "      <td>13.413555</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.985239</td>\n",
       "      <td>0.859911</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Treino        Teste\n",
       "MAE         23.950637    41.683038\n",
       "MSE        962.984248  8140.760631\n",
       "RMSE        31.031987    90.226164\n",
       "MAPE        36.381970    13.413555\n",
       "R-squared    0.985239     0.859911"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "RandomForestRegressor\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>11.574877</td>\n",
       "      <td>36.681784</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>398.139804</td>\n",
       "      <td>7158.794449</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>19.953441</td>\n",
       "      <td>84.609659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>30.578551</td>\n",
       "      <td>8.370187</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.993897</td>\n",
       "      <td>0.876809</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Treino        Teste\n",
       "MAE         11.574877    36.681784\n",
       "MSE        398.139804  7158.794449\n",
       "RMSE        19.953441    84.609659\n",
       "MAPE        30.578551     8.370187\n",
       "R-squared    0.993897     0.876809"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "PCARegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>53.294235</td>\n",
       "      <td>69.551406</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>5604.668222</td>\n",
       "      <td>13437.109407</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>74.864332</td>\n",
       "      <td>115.918546</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>249.648657</td>\n",
       "      <td>22.293332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.914090</td>\n",
       "      <td>0.768769</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Treino         Teste\n",
       "MAE          53.294235     69.551406\n",
       "MSE        5604.668222  13437.109407\n",
       "RMSE         74.864332    115.918546\n",
       "MAPE        249.648657     22.293332\n",
       "R-squared     0.914090      0.768769"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "PolyRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>14.071769</td>\n",
       "      <td>31.924466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>324.874632</td>\n",
       "      <td>5085.567821</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>18.024279</td>\n",
       "      <td>71.313167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>20.140633</td>\n",
       "      <td>7.897775</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.995020</td>\n",
       "      <td>0.912486</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Treino        Teste\n",
       "MAE         14.071769    31.924466\n",
       "MSE        324.874632  5085.567821\n",
       "RMSE        18.024279    71.313167\n",
       "MAPE        20.140633     7.897775\n",
       "R-squared    0.995020     0.912486"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n"
     ]
    }
   ],
   "source": [
    "run_all_algs(X_train, \n",
    "             y_train,\n",
    "             X_test,\n",
    "             y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "804b778a-2b42-4c54-8b1c-bd5762955405",
   "metadata": {},
   "source": [
    "# Prevendo para as máscaras geradas pela a unet - Sem normalização"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "29431178-7ffa-4e31-b383-d301d11c4d20",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "GradientBoostingRegressor\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>5.753149</td>\n",
       "      <td>41.883434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>53.400895</td>\n",
       "      <td>7405.489313</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>7.307592</td>\n",
       "      <td>86.055153</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>6.286146</td>\n",
       "      <td>10.472568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.999181</td>\n",
       "      <td>0.872563</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Treino        Teste\n",
       "MAE         5.753149    41.883434\n",
       "MSE        53.400895  7405.489313\n",
       "RMSE        7.307592    86.055153\n",
       "MAPE        6.286146    10.472568\n",
       "R-squared   0.999181     0.872563"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "LinearRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>23.974582</td>\n",
       "      <td>59.531243</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>962.880129</td>\n",
       "      <td>9437.620208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>31.030310</td>\n",
       "      <td>97.147415</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>37.276645</td>\n",
       "      <td>19.039013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.985241</td>\n",
       "      <td>0.837594</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Treino        Teste\n",
       "MAE         23.974582    59.531243\n",
       "MSE        962.880129  9437.620208\n",
       "RMSE        31.030310    97.147415\n",
       "MAPE        37.276645    19.039013\n",
       "R-squared    0.985241     0.837594"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "LassoRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>26.431473</td>\n",
       "      <td>57.662196</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>1171.437241</td>\n",
       "      <td>9605.645580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>34.226265</td>\n",
       "      <td>98.008395</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>27.699546</td>\n",
       "      <td>18.401953</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.982044</td>\n",
       "      <td>0.834702</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Treino        Teste\n",
       "MAE          26.431473    57.662196\n",
       "MSE        1171.437241  9605.645580\n",
       "RMSE         34.226265    98.008395\n",
       "MAPE         27.699546    18.401953\n",
       "R-squared     0.982044     0.834702"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "RidgeRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>23.950637</td>\n",
       "      <td>59.496301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>962.984248</td>\n",
       "      <td>9414.481077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>31.031987</td>\n",
       "      <td>97.028249</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>36.381970</td>\n",
       "      <td>18.984062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.985239</td>\n",
       "      <td>0.837992</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Treino        Teste\n",
       "MAE         23.950637    59.496301\n",
       "MSE        962.984248  9414.481077\n",
       "RMSE        31.031987    97.028249\n",
       "MAPE        36.381970    18.984062\n",
       "R-squared    0.985239     0.837992"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "RandomForestRegressor\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>11.003918</td>\n",
       "      <td>43.362838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>330.879983</td>\n",
       "      <td>7778.158405</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>18.190107</td>\n",
       "      <td>88.193868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>34.847223</td>\n",
       "      <td>10.396883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.994928</td>\n",
       "      <td>0.866150</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Treino        Teste\n",
       "MAE         11.003918    43.362838\n",
       "MSE        330.879983  7778.158405\n",
       "RMSE        18.190107    88.193868\n",
       "MAPE        34.847223    10.396883\n",
       "R-squared    0.994928     0.866150"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "PCARegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>53.294235</td>\n",
       "      <td>70.666161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>5604.668222</td>\n",
       "      <td>13343.222121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>74.864332</td>\n",
       "      <td>115.512866</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>249.648657</td>\n",
       "      <td>23.037587</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.914090</td>\n",
       "      <td>0.770385</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Treino         Teste\n",
       "MAE          53.294235     70.666161\n",
       "MSE        5604.668222  13343.222121\n",
       "RMSE         74.864332    115.512866\n",
       "MAPE        249.648657     23.037587\n",
       "R-squared     0.914090      0.770385"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "PolyRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>14.071769</td>\n",
       "      <td>40.712081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>324.874632</td>\n",
       "      <td>5728.659042</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>18.024279</td>\n",
       "      <td>75.687906</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>20.140633</td>\n",
       "      <td>10.664158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.995020</td>\n",
       "      <td>0.901419</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Treino        Teste\n",
       "MAE         14.071769    40.712081\n",
       "MSE        324.874632  5728.659042\n",
       "RMSE        18.024279    75.687906\n",
       "MAPE        20.140633    10.664158\n",
       "R-squared    0.995020     0.901419"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n"
     ]
    }
   ],
   "source": [
    "run_all_algs(X_train, \n",
    "             y_train,\n",
    "             unet_testing_features.values,\n",
    "             y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d51c5dd7-2f3f-41f4-bc55-860f729a1d6f",
   "metadata": {},
   "source": [
    "# Prevendo os pesos para as márcas previstas pela unet | Normalizado"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "8a311f27-182e-4b6e-8823-c3da02d87ea1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "GradientBoostingRegressor\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>5.753149</td>\n",
       "      <td>41.296664</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>53.400895</td>\n",
       "      <td>7347.096177</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>7.307592</td>\n",
       "      <td>85.715204</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>6.286146</td>\n",
       "      <td>10.406900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.999181</td>\n",
       "      <td>0.873568</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Treino        Teste\n",
       "MAE         5.753149    41.296664\n",
       "MSE        53.400895  7347.096177\n",
       "RMSE        7.307592    85.715204\n",
       "MAPE        6.286146    10.406900\n",
       "R-squared   0.999181     0.873568"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "LinearRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>23.975293</td>\n",
       "      <td>59.529593</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>962.880039</td>\n",
       "      <td>9437.755647</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>31.030308</td>\n",
       "      <td>97.148112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>37.311182</td>\n",
       "      <td>19.038821</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.985241</td>\n",
       "      <td>0.837591</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Treino        Teste\n",
       "MAE         23.975293    59.529593\n",
       "MSE        962.880039  9437.755647\n",
       "RMSE        31.030308    97.148112\n",
       "MAPE        37.311182    19.038821\n",
       "R-squared    0.985241     0.837591"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "LassoRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>31.956953</td>\n",
       "      <td>55.031958</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>1804.489268</td>\n",
       "      <td>8736.306219</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>42.479280</td>\n",
       "      <td>93.468210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>21.587117</td>\n",
       "      <td>17.528081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.972340</td>\n",
       "      <td>0.849662</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Treino        Teste\n",
       "MAE          31.956953    55.031958\n",
       "MSE        1804.489268  8736.306219\n",
       "RMSE         42.479280    93.468210\n",
       "MAPE         21.587117    17.528081\n",
       "R-squared     0.972340     0.849662"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "RidgeRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>28.597508</td>\n",
       "      <td>55.451863</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>1448.371179</td>\n",
       "      <td>8942.838852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>38.057472</td>\n",
       "      <td>94.566584</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>15.823772</td>\n",
       "      <td>17.916019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.977799</td>\n",
       "      <td>0.846108</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Treino        Teste\n",
       "MAE          28.597508    55.451863\n",
       "MSE        1448.371179  8942.838852\n",
       "RMSE         38.057472    94.566584\n",
       "MAPE         15.823772    17.916019\n",
       "R-squared     0.977799     0.846108"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "RandomForestRegressor\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>11.733789</td>\n",
       "      <td>43.153730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>407.688055</td>\n",
       "      <td>7759.390649</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>20.191287</td>\n",
       "      <td>88.087403</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>38.297140</td>\n",
       "      <td>10.327579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.993751</td>\n",
       "      <td>0.866473</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Treino        Teste\n",
       "MAE         11.733789    43.153730\n",
       "MSE        407.688055  7759.390649\n",
       "RMSE        20.191287    88.087403\n",
       "MAPE        38.297140    10.327579\n",
       "R-squared    0.993751     0.866473"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "PCARegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>37.241035</td>\n",
       "      <td>54.234824</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>2470.607808</td>\n",
       "      <td>9209.688890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>49.705209</td>\n",
       "      <td>95.967124</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>92.887716</td>\n",
       "      <td>17.147199</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.962130</td>\n",
       "      <td>0.841516</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Treino        Teste\n",
       "MAE          37.241035    54.234824\n",
       "MSE        2470.607808  9209.688890\n",
       "RMSE         49.705209    95.967124\n",
       "MAPE         92.887716    17.147199\n",
       "R-squared     0.962130     0.841516"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n",
      "\n",
      "PolyRegression\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>Treino</th>\n",
       "      <th>Teste</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>19.584782</td>\n",
       "      <td>49.269408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>622.533633</td>\n",
       "      <td>10944.823711</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>24.950624</td>\n",
       "      <td>104.617511</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>28.739563</td>\n",
       "      <td>19.989758</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.990458</td>\n",
       "      <td>0.811657</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Treino         Teste\n",
       "MAE         19.584782     49.269408\n",
       "MSE        622.533633  10944.823711\n",
       "RMSE        24.950624    104.617511\n",
       "MAPE        28.739563     19.989758\n",
       "R-squared    0.990458      0.811657"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________\n"
     ]
    }
   ],
   "source": [
    "run_all_algs(X_train_scaled, \n",
    "             y_train,\n",
    "             Unet_X_test_scaled,\n",
    "             y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ca8dabab-55f3-40a0-87cc-7713ba15659e",
   "metadata": {},
   "source": [
    "# KFold Base Completa"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "84164fb3-1016-45c5-b3da-5fb5ff5040bc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[GradientBoostingRegressor]\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>Média</th>\n",
       "      <th>Desvio Padrão</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.948509</td>\n",
       "      <td>0.046907</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>30.191133</td>\n",
       "      <td>13.422909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>4231.828935</td>\n",
       "      <td>5360.343515</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>53.645025</td>\n",
       "      <td>36.797285</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>0.431264</td>\n",
       "      <td>1.102556</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Média  Desvio Padrão\n",
       "R-squared     0.948509       0.046907\n",
       "MAE          30.191133      13.422909\n",
       "MSE        4231.828935    5360.343515\n",
       "RMSE         53.645025      36.797285\n",
       "MAPE          0.431264       1.102556"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LinearRegression]\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>Média</th>\n",
       "      <th>Desvio Padrão</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.916003</td>\n",
       "      <td>0.067720</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>38.634638</td>\n",
       "      <td>19.924548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>7165.200820</td>\n",
       "      <td>8997.645655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>69.881547</td>\n",
       "      <td>47.767878</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>0.403067</td>\n",
       "      <td>0.975666</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Média  Desvio Padrão\n",
       "R-squared     0.916003       0.067720\n",
       "MAE          38.634638      19.924548\n",
       "MSE        7165.200820    8997.645655\n",
       "RMSE         69.881547      47.767878\n",
       "MAPE          0.403067       0.975666"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LassoRegression]\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>Média</th>\n",
       "      <th>Desvio Padrão</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.948340</td>\n",
       "      <td>0.052341</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>33.058583</td>\n",
       "      <td>12.211045</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>3806.849370</td>\n",
       "      <td>5773.404311</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>51.553907</td>\n",
       "      <td>33.897552</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>0.599363</td>\n",
       "      <td>1.556591</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Média  Desvio Padrão\n",
       "R-squared     0.948340       0.052341\n",
       "MAE          33.058583      12.211045\n",
       "MSE        3806.849370    5773.404311\n",
       "RMSE         51.553907      33.897552\n",
       "MAPE          0.599363       1.556591"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[RidgeRegression]\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>Média</th>\n",
       "      <th>Desvio Padrão</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.916283</td>\n",
       "      <td>0.067658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>38.586087</td>\n",
       "      <td>19.928811</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>7142.836878</td>\n",
       "      <td>8983.591178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>69.750226</td>\n",
       "      <td>47.725704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>0.399285</td>\n",
       "      <td>0.964782</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Média  Desvio Padrão\n",
       "R-squared     0.916283       0.067658\n",
       "MAE          38.586087      19.928811\n",
       "MSE        7142.836878    8983.591178\n",
       "RMSE         69.750226      47.725704\n",
       "MAPE          0.399285       0.964782"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[RandomForestRegressor]\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>Média</th>\n",
       "      <th>Desvio Padrão</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.945702</td>\n",
       "      <td>0.044853</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>31.876821</td>\n",
       "      <td>13.012178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>4243.634894</td>\n",
       "      <td>5243.925198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>54.939476</td>\n",
       "      <td>35.004126</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>0.512901</td>\n",
       "      <td>1.338996</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Média  Desvio Padrão\n",
       "R-squared     0.945702       0.044853\n",
       "MAE          31.876821      13.012178\n",
       "MSE        4243.634894    5243.925198\n",
       "RMSE         54.939476      35.004126\n",
       "MAPE          0.512901       1.338996"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[PCARegression]\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>Média</th>\n",
       "      <th>Desvio Padrão</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.863217</td>\n",
       "      <td>0.063852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>59.450182</td>\n",
       "      <td>16.137315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>8314.969495</td>\n",
       "      <td>8029.216407</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>84.182258</td>\n",
       "      <td>35.047352</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>1.735975</td>\n",
       "      <td>4.667051</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Média  Desvio Padrão\n",
       "R-squared     0.863217       0.063852\n",
       "MAE          59.450182      16.137315\n",
       "MSE        8314.969495    8029.216407\n",
       "RMSE         84.182258      35.047352\n",
       "MAPE          1.735975       4.667051"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[PolyRegression]\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>Média</th>\n",
       "      <th>Desvio Padrão</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.926055</td>\n",
       "      <td>0.049774</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>34.117808</td>\n",
       "      <td>13.069283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>5110.334955</td>\n",
       "      <td>5444.705029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>62.247616</td>\n",
       "      <td>35.150665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>0.377524</td>\n",
       "      <td>0.920124</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Média  Desvio Padrão\n",
       "R-squared     0.926055       0.049774\n",
       "MAE          34.117808      13.069283\n",
       "MSE        5110.334955    5444.705029\n",
       "RMSE         62.247616      35.150665\n",
       "MAPE          0.377524       0.920124"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ApplyesKFoldV2(X.values, y.values, k=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fdf4a294-b9a6-4745-9f21-db79bac6c220",
   "metadata": {},
   "source": [
    "# PCA + Kfold Base Completa"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "61eb02f9-8725-4136-af94-1ccd4287cfcc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Para preservar 90% da variance é necessário 2\n",
      "Ratio:  [0.89758847 0.0477517 ]\n",
      "[GradientBoostingRegressor]\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>Média</th>\n",
       "      <th>Desvio Padrão</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.927278</td>\n",
       "      <td>0.042374</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>41.510625</td>\n",
       "      <td>15.080491</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>4957.346326</td>\n",
       "      <td>5501.620498</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>62.034263</td>\n",
       "      <td>33.303100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>0.515546</td>\n",
       "      <td>1.273078</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Média  Desvio Padrão\n",
       "R-squared     0.927278       0.042374\n",
       "MAE          41.510625      15.080491\n",
       "MSE        4957.346326    5501.620498\n",
       "RMSE         62.034263      33.303100\n",
       "MAPE          0.515546       1.273078"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LinearRegression]\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>Média</th>\n",
       "      <th>Desvio Padrão</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.919737</td>\n",
       "      <td>0.059523</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>42.418198</td>\n",
       "      <td>15.129838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>5634.693078</td>\n",
       "      <td>7179.097503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>64.748453</td>\n",
       "      <td>37.978031</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>0.218223</td>\n",
       "      <td>0.299254</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Média  Desvio Padrão\n",
       "R-squared     0.919737       0.059523\n",
       "MAE          42.418198      15.129838\n",
       "MSE        5634.693078    7179.097503\n",
       "RMSE         64.748453      37.978031\n",
       "MAPE          0.218223       0.299254"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LassoRegression]\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>Média</th>\n",
       "      <th>Desvio Padrão</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.920902</td>\n",
       "      <td>0.058866</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>42.300635</td>\n",
       "      <td>14.984307</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>5514.065433</td>\n",
       "      <td>7039.656405</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>64.209033</td>\n",
       "      <td>37.299672</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>0.264461</td>\n",
       "      <td>0.439303</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Média  Desvio Padrão\n",
       "R-squared     0.920902       0.058866\n",
       "MAE          42.300635      14.984307\n",
       "MSE        5514.065433    7039.656405\n",
       "RMSE         64.209033      37.299672\n",
       "MAPE          0.264461       0.439303"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[RidgeRegression]\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>Média</th>\n",
       "      <th>Desvio Padrão</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.919794</td>\n",
       "      <td>0.059497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>42.410597</td>\n",
       "      <td>15.118030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>5629.132130</td>\n",
       "      <td>7173.181825</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>64.722452</td>\n",
       "      <td>37.949128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>0.218863</td>\n",
       "      <td>0.301357</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Média  Desvio Padrão\n",
       "R-squared     0.919794       0.059497\n",
       "MAE          42.410597      15.118030\n",
       "MSE        5629.132130    7173.181825\n",
       "RMSE         64.722452      37.949128\n",
       "MAPE          0.218863       0.301357"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[RandomForestRegressor]\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>Média</th>\n",
       "      <th>Desvio Padrão</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.930707</td>\n",
       "      <td>0.037181</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>40.617042</td>\n",
       "      <td>14.757823</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>4774.936163</td>\n",
       "      <td>4981.120872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>61.544826</td>\n",
       "      <td>31.419271</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>0.504724</td>\n",
       "      <td>1.246319</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Média  Desvio Padrão\n",
       "R-squared     0.930707       0.037181\n",
       "MAE          40.617042      14.757823\n",
       "MSE        4774.936163    4981.120872\n",
       "RMSE         61.544826      31.419271\n",
       "MAPE          0.504724       1.246319"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[PCARegression]\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>Média</th>\n",
       "      <th>Desvio Padrão</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.919737</td>\n",
       "      <td>0.059523</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>42.418198</td>\n",
       "      <td>15.129838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>5634.693078</td>\n",
       "      <td>7179.097503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>64.748453</td>\n",
       "      <td>37.978031</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>0.218223</td>\n",
       "      <td>0.299254</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Média  Desvio Padrão\n",
       "R-squared     0.919737       0.059523\n",
       "MAE          42.418198      15.129838\n",
       "MSE        5634.693078    7179.097503\n",
       "RMSE         64.748453      37.978031\n",
       "MAPE          0.218223       0.299254"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[PolyRegression]\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>Média</th>\n",
       "      <th>Desvio Padrão</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>R-squared</th>\n",
       "      <td>0.799249</td>\n",
       "      <td>0.402122</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAE</th>\n",
       "      <td>47.659149</td>\n",
       "      <td>31.904285</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSE</th>\n",
       "      <td>20897.591240</td>\n",
       "      <td>49456.644469</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSE</th>\n",
       "      <td>92.247541</td>\n",
       "      <td>111.301314</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAPE</th>\n",
       "      <td>0.195444</td>\n",
       "      <td>0.306810</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  Média  Desvio Padrão\n",
       "R-squared      0.799249       0.402122\n",
       "MAE           47.659149      31.904285\n",
       "MSE        20897.591240   49456.644469\n",
       "RMSE          92.247541     111.301314\n",
       "MAPE           0.195444       0.306810"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_pca = apply_pca_reduce(X.values, 2)\n",
    "\n",
    "ApplyesKFoldV2(X_pca, y.values, k=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3dbcfc4-d0a4-47d5-9084-57d8027428f3",
   "metadata": {},
   "source": [
    "## Polinomial Features -  Qualquer  regressor pode ser adicionado ao pipeline\n",
    "\n",
    "#### Obs: Não executar transformação de features\n",
    "\n",
    "\n",
    "Foi selecionado um dos modelos de regressão que performaram bem para visualizar dua predições"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "455e9bfc-a90d-4a72-8d25-87076e9a560c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "               Treino\n",
      "MAE         14.051310\n",
      "MSE        324.694800\n",
      "RMSE        18.019290\n",
      "MAPE        21.042099\n",
      "R-squared    0.995023\n",
      "                 Teste\n",
      "MAE          32.486902\n",
      "MSE        5132.968023\n",
      "RMSE         71.644735\n",
      "MAPE          8.406508\n",
      "R-squared     0.911670\n"
     ]
    }
   ],
   "source": [
    "model = make_pipeline(PolynomialFeatures(degree=3), Lasso(0.5)).fit(X_train, y_train)\n",
    "\n",
    "print_evaluate(y_train, model.predict(X_train), \"Treino\")\n",
    "print_evaluate(y_test, model.predict(X_test), \"Teste\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7dd8588a-d8c3-4982-9224-f19456f5758b",
   "metadata": {},
   "source": [
    "# Figuras para explicar o modelo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "fbcf08db-79f4-4359-8c1c-a06866fe8df2",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = make_pipeline(PolynomialFeatures(degree=3), Lasso(0.5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "id": "deb8bdc8-75e9-42bd-8acf-9691d575f3b1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# fig, ax = plt.subplots(figsize=(6, 10))\n",
    "fig, axs = plt.subplots(nrows=1, ncols=2, figsize=(12, 5))\n",
    "\n",
    "\n",
    "def make_plot(ax, model, x, y, title):\n",
    "\n",
    "    model = PredictionError(model, ax=ax).fit(X_train, y_train)\n",
    "    score = model.score(x, y)\n",
    "    model.finalize()\n",
    "\n",
    "    ax.set_xlabel(\"(y) Peso Predito - gramas\")\n",
    "    ax.set_ylabel(\"(ŷ) Peso Real - gramas\")\n",
    "    ax.legend([f\"R² = {round(score, 3)}\", \"Melhor Ajuste\", \"Identidade\"])\n",
    "    ax.set_title(f\"Análise de Erro de Predição - {title}\")\n",
    "\n",
    "    ax.grid(color='gray', \n",
    "            linestyle='--',\n",
    "            linewidth=0.5,\n",
    "            alpha=0.5)\n",
    "    \n",
    "    return score\n",
    "\n",
    "test_score = make_plot(axs[1], \n",
    "                       model,\n",
    "                       X_test,\n",
    "                       y_test, \n",
    "                       title=\"Teste\")\n",
    "\n",
    "train_score = make_plot(axs[0],\n",
    "                        model,\n",
    "                        X_train,\n",
    "                        y_train, \n",
    "                        title=\"Treino\")\n",
    "\n",
    "plt.savefig(\"analise_erro_predicao.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "id": "6b5d1a57-042d-4600-9503-326242b16abe",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "visualizer = ResidualsPlot(model, hist=False).fit(X_train, y_train)\n",
    "\n",
    "test_score = visualizer.score(X_test, y_test)\n",
    "\n",
    "visualizer.finalize()\n",
    "\n",
    "visualizer.ax.legend([f\"Treino R² = {round(train_score, 3)}\", f\"Teste R² = {round(test_score, 3)}\"], loc=\"upper left\")\n",
    "\n",
    "visualizer.ax.set_xlabel(\"(y) Peso Predito - gramas\")\n",
    "visualizer.ax.set_ylabel(\"Resíduos\")\n",
    "visualizer.ax.set_title(\"Análise Residual do Modelo de Regressão\")\n",
    "visualizer.ax.grid(color='gray', \n",
    "                   linestyle='--',\n",
    "                   linewidth=0.5,\n",
    "                   alpha=0.5)\n",
    "\n",
    "plt.savefig(\"analise_residuos_predicao.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d764180c-04ac-4b5e-829c-46cf9f60f176",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}