model.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1. Data Load"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os\n",
    "\n",
    "# NOTE : data load\n",
    "path = os.getcwd()\n",
    "train = pd.read_csv(f'{path}/open/train.csv')\n",
    "test = pd.read_csv(f'{path}/open/test.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. Preprocessing  \n",
    "<br>\n",
    "<br>\n",
    "\n",
    "##### 1) feature engineering\n",
    "##### 2) drop columns\n",
    "##### 3) bining(변수 구간화)\n",
    "##### 4) encoding\n",
    "##### 5) scaling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from module import pre_1_datetime, pre_2_weekday, pre_3_ARI_PORT\n",
    "from module import pre_4_ratio_under_water, pre_5_ratio_under_water_DEADWEIGHT\n",
    "from module import pre_6_dist_binding, pre_7_drop, pre_8_encoding\n",
    "\n",
    "\n",
    "# feature engineering / drop columns / bining\n",
    "def preprocessing_1(data):\n",
    "    \n",
    "    data = pre_1_datetime(data)\n",
    "    data = pre_2_weekday(data)\n",
    "    data = pre_3_ARI_PORT(data)\n",
    "    data = pre_4_ratio_under_water(data)\n",
    "    data = pre_5_ratio_under_water_DEADWEIGHT(data)\n",
    "    data = pre_6_dist_binding(data)\n",
    "    data = pre_7_drop(data)\n",
    "    \n",
    "    return data\n",
    "\n",
    "\n",
    "# encoding\n",
    "def preprocessing_2(train, test):\n",
    "\n",
    "    train, test = pre_8_encoding(train, test)\n",
    "\n",
    "    return train, test\n",
    "\n",
    "\n",
    "train = preprocessing_1(train)\n",
    "test = preprocessing_1(test)\n",
    "\n",
    "train, test = preprocessing_2(train, test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DIST = 0(train) : (157293, 28)\n",
      "DIST != 0(train) : (234646, 28)\n",
      "\n",
      "✅ 전처리 완료\n"
     ]
    }
   ],
   "source": [
    "from module import model_1_split_DIST, modeling_2_creterion_KF, modeling_3_make_creterion\n",
    "from module import pre_9_drop, pre_10_scaling\n",
    "\n",
    "\n",
    "\n",
    "# NOTE : Modeling 1 / 데이터셋 분할\n",
    "# DIST가 0인 samples, 0이 아닌 samples로 분할\n",
    "# 분할이유 \n",
    "    #: EDA결과 DIST가 0인 경우 대부분 target(CI_HOUR)가 0이였음.\n",
    "    #  따라서 DIST가 0이 아닌 경우는 DIST가 0인 상황을 제외하고 학습하는게 좋다고 판단.\n",
    "train_0, train_1, test_0, test_1 = model_1_split_DIST(train, test)\n",
    "\n",
    "\n",
    "\n",
    "# NOTE : Modeling 2 / K-Fold 데이터셋 분할 기준 생성(1)\n",
    "# train_0, test_0 : target(CI_HOUR)이 24시간 미만이면 0, 이상이면 1\n",
    "# train_1, test_1 : target(CI_HOUR)이 100시간 미만이면 0, 이상이면 1\n",
    "train_0 = modeling_2_creterion_KF(train_0, 24)\n",
    "train_1 = modeling_2_creterion_KF(train_1, 100)\n",
    "\n",
    "\n",
    "\n",
    "# DIST가 0인 데이터셋에 한해서 'weekday'column 제거\n",
    "train_0 = pre_9_drop(train_0)\n",
    "test_0 = pre_9_drop(test_0)\n",
    "\n",
    "\n",
    "\n",
    "# stratified KF 검증 시 데이터셋 분할 기준 생성\n",
    "SPLIT_CRETERION_0, SPLIT_CRETERION_1 = modeling_3_make_creterion(train_0, train_1)\n",
    "\n",
    "\n",
    "# scaling\n",
    "train_x_0, train_y_0, test_0 = pre_10_scaling(train_0, test_0)\n",
    "train_x_1, train_y_1, test_1 = pre_10_scaling(train_1, test_1)\n",
    "\n",
    "print('\\n✅ 전처리 완료')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. Modeling  \n",
    "<br>\n",
    "<br>\n",
    "\n",
    "##### 1) feature importance 계산\n",
    "##### 2) drop columns\n",
    "##### 3) best model 찾기\n",
    "##### 4) stratified K-Fold 검증 및 추론"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from module import modeling_4_feature_importance, modeling_5_drop_columns_based_fi\n",
    "# from module import modeling_6_grid_search, modeling_7_train_evaluation\n",
    "# from xgboost import XGBRegressor\n",
    "# import warnings\n",
    "\n",
    "# warnings.filterwarnings(action='ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# def modeling(train_x, train_y, test, model, model_name, threshold, params_grid, SPLIT_CRETERION):\n",
    "\n",
    "\n",
    "#     # NOTE : Modeling 4 / 변수 중요도 추출 및 시각화\n",
    "#     trained_model, feature_importances = modeling_4_feature_importance(model, model_name, train_x, train_y)\n",
    "\n",
    "\n",
    "#     # NOTE : Modeling 5 / 변수 중요도 기반 전처리\n",
    "#     # threshold보다 중요도가 낮은 column은 제거\n",
    "#     # 실험에 의해 threshold = 0.004가 가장 좋은 성능을 보여주는 것 확인\n",
    "#     train_x, test = modeling_5_drop_columns_based_fi(train_x, test, feature_importances, threshold)\n",
    "\n",
    "\n",
    "\n",
    "#     # NOTE : Modeling 6 / 최적 하이퍼파라미터\n",
    "#     # Grid Search를 사용해 모델의 최적 하이퍼파라미터 조합발견\n",
    "#     best_model, best_params = modeling_6_grid_search(train_x, train_y, model, params_grid)\n",
    "#     print('\\n\\n✅ Best Hyperparameters :\\n', best_params)\n",
    "\n",
    "\n",
    "#     # NOTE : Modeling 7 / Stratified K-Fold\n",
    "#     # best model에 Stratified K-Fold적용 후 학습, 검증, 추론 수행\n",
    "#     final_predictions = modeling_7_train_evaluation(train_x, train_y, test, best_model, SPLIT_CRETERION)\n",
    "\n",
    "#     return final_predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from module import modeling_4_feature_importance, modeling_5_grid_search, modeling_6_train_evaluation\n",
    "#from module import modeling_6_grid_search, modeling_7_train_evaluation\n",
    "from xgboost import XGBRegressor\n",
    "import warnings\n",
    "\n",
    "warnings.filterwarnings(action='ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def modeling(train_x, train_y, test, model, model_name, threshold, params_grid, SPLIT_CRETERION):\n",
    "\n",
    "\n",
    "    # NOTE : Modeling 4 / 변수 중요도 추출 및 시각화, 시각화\n",
    "    # threshold보다 중요도가 낮은 column은 제거\n",
    "    # 실험에 의해 threshold = 0.004가 가장 좋은 성능을 보여주는 것 확인\n",
    "    train_x, test, trained_model, feature_importances = modeling_4_feature_importance(train_x, train_y, test, model, model_name, threshold)\n",
    "\n",
    "\n",
    "    # NOTE : Modeling 6 / 최적 하이퍼파라미터\n",
    "    # Grid Search를 사용해 모델의 최적 하이퍼파라미터 조합발견\n",
    "    best_model, best_params = modeling_5_grid_search(train_x, train_y, model, params_grid)\n",
    "    print('\\n\\n✅ Best Hyperparameters :\\n', best_params)\n",
    "\n",
    "\n",
    "    # NOTE : Modeling 7 / Stratified K-Fold\n",
    "    # best model에 Stratified K-Fold적용 후 학습, 검증, 추론 수행\n",
    "    final_predictions = modeling_6_train_evaluation(train_x, train_y, test, best_model, SPLIT_CRETERION)\n",
    "\n",
    "    return final_predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1) DIST가 0인 경우"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "params_grid = {\n",
    "    'eta': [0.02, 0.03, 0.04],  # 학습률 (learning rate)\n",
    "    'n_estimators': range(400, 550, 50),  # 예측기(트리)의 개수\n",
    "    'min_child_weight': [1, 1.25, 1.5],  # 리프 노드의 최소 가중치 합    \n",
    "    'max_depth': range(10, 16, 2),  # 트리의 최대 깊이    \n",
    "    'lambda': [2.5, 3, 3.5],  # L2 규제 (람다)\n",
    "    'alpha': [ 3 ],  # L1 규제 (알파)\n",
    "\n",
    "    'gamma': [0.2],  # 트리 노드를 분할하는 데 필요한 최소 손실 감소\n",
    "    'colsample_bytree': [0.8],  # 트리를 구성할 때 열 샘플링 비율\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Tune for XGBRegressor.\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "💡 제거된 column들 \n",
      ": ['DIST', 'BUILT', 'DEADWEIGHT', 'DEPTH', 'LENGTH', 'year', 'ratio_under_water', 'ratio_under_water_DEAD', 'DIST_binded']\n",
      "\n",
      "\n",
      "✅ Best Hyperparameters :\n",
      " {'alpha': 3, 'colsample_bytree': 0.8, 'eta': 0.04, 'gamma': 0.2, 'lambda': 2.5, 'max_depth': 10, 'min_child_weight': 1, 'n_estimators': 400}\n",
      "\n",
      "\n",
      "Prediction Score\n",
      "Validation : MAE for each fold: [0.011797943961677774, 0.024097228371963663, 0.002388787113794518, 0.021294653029057357, 0.025968146967173335]\n",
      "Validation : MAE: 0.01710935188873333\n"
     ]
    }
   ],
   "source": [
    "model_0 = XGBRegressor()\n",
    "model_name_0 = 'XGBRegressor'\n",
    "\n",
    "\n",
    "\n",
    "final_predictions_0 = modeling(train_x_0, train_y_0, test_0, model_0, model_name_0, 0.004, params_grid, SPLIT_CRETERION_0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2) DIST가 0보다 큰 경우"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Tune for XGBRegressor.\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "💡 제거된 column들 \n",
      ": ['BUILT', 'DEPTH', 'DRAUGHT', 'U_WIND', 'V_WIND', 'AIR_TEMPERATURE', 'BN', 'ATA_LT', 'month', 'day', 'minute', 'weekday_binded', 'ratio_under_water']\n",
      "\n",
      "\n",
      "✅ Best Hyperparameters :\n",
      " {'alpha': 3, 'colsample_bytree': 0.8, 'eta': 0.02, 'gamma': 0.2, 'lambda': 3, 'max_depth': 14, 'min_child_weight': 1.25, 'n_estimators': 450}\n",
      "\n",
      "\n",
      "Prediction Score\n",
      "Validation : MAE for each fold: [73.27557026392293, 73.70083976964337, 74.01356542101907, 73.82304362388375, 72.77460029650194]\n",
      "Validation : MAE: 73.51752387499421\n"
     ]
    }
   ],
   "source": [
    "model_1 = XGBRegressor()\n",
    "model_name_1 = 'XGBRegressor'\n",
    "\n",
    "\n",
    "\n",
    "final_predictions_1 = modeling(train_x_1, train_y_1, test_1, model_1, model_name_1, 0.012, params_grid, SPLIT_CRETERION_1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. Inference"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "test = pd.read_csv(f'{path}/open/test.csv')\n",
    "\n",
    "test_0 = test[test['DIST'] == 0]\n",
    "test_1 = test[test['DIST'] != 0]\n",
    "\n",
    "test_0['CI_HOUR'] = final_predictions_0\n",
    "test_1['CI_HOUR'] = final_predictions_1\n",
    "\n",
    "\n",
    "# SAMPLE_ID 순으로 정렬 후 submit을 위한 column만 남기기\n",
    "submit = pd.concat([test_0, test_1]).sort_values('SAMPLE_ID')[['SAMPLE_ID', 'CI_HOUR']]\n",
    "\n",
    "# save\n",
    "submit.to_csv(f'{path}/open/submit.csv', index=False)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "HD",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}