research/04_model_trainer.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## step 1: go to the right folder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'/Users/krystalli/Documents/End-to-End-Wine-Quality-Prediction/research'"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import os\n",
    "%pwd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'/Users/krystalli/Documents/End-to-End-Wine-Quality-Prediction'"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "os.chdir('../')\n",
    "%pwd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## step 2: update config.yaml"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## step 3: update params.yaml\n",
    "\n",
    "Since I used ElasticNet for modelling training, alpha and li_ratio parameters are what need to pay attention to bcz they are the regularization parameters. (ElasticNet: https://scikit-learn.org/dev/modules/generated/sklearn.linear_model.ElasticNet.html)\n",
    "\n",
    "If I want to change these two parameters in the future to improve model performance, can simple change them in the params.yaml\n",
    "\n",
    "I can also use other models like linear regression/random forest/XGBoost, and hyper parameter tuning, simple write the parameters in the params.yaml the way I did for ElasticNet."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 4: update entity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "from dataclasses import dataclass\n",
    "from pathlib import Path\n",
    "\n",
    "# initialize all the variables\n",
    "@dataclass(frozen=True)\n",
    "class ModelTrainerConfig:\n",
    "    root_dir: Path\n",
    "    train_data_path: Path\n",
    "    test_data_path: Path\n",
    "    model_name: str\n",
    "    alpha: float\n",
    "    l1_ratio: float\n",
    "    target_column: str"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 5: update configuration manager"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "from wine_quality_prediction_project.constants import *\n",
    "from wine_quality_prediction_project.utils.common import read_yaml, create_directories\n",
    "\n",
    "class ConfigurationManager:\n",
    "    def __init__(\n",
    "        self,\n",
    "        config_filepath = CONFIG_FILE_PATH,\n",
    "        params_filepath = PARAMS_FILE_PATH,\n",
    "        schema_filepath = SCHEMA_FILE_PATH):\n",
    "\n",
    "        self.config = read_yaml(config_filepath)\n",
    "        self.params = read_yaml(params_filepath)\n",
    "        self.schema = read_yaml(schema_filepath)\n",
    "\n",
    "        create_directories([self.config.artifacts_root])\n",
    "        \n",
    "    def get_model_trainer_config(self) -> ModelTrainerConfig:\n",
    "        config = self.config.model_trainer\n",
    "        params = self.params.ElasticNet\n",
    "        schema = self.schema.TARGET_COLUMN\n",
    "        \n",
    "        create_directories([config.root_dir])\n",
    "        \n",
    "        model_trainer_config = ModelTrainerConfig(\n",
    "            root_dir=config.root_dir,\n",
    "            train_data_path=config.train_data_path,\n",
    "            test_data_path=config.test_data_path,\n",
    "            model_name=config.model_name,\n",
    "            alpha=params.alpha,\n",
    "            l1_ratio=params.l1_ratio,\n",
    "            target_column=schema.name\n",
    "        )\n",
    "        \n",
    "        return model_trainer_config"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 6: update components"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import os\n",
    "from wine_quality_prediction_project import logger\n",
    "from sklearn.linear_model import ElasticNet\n",
    "import joblib\n",
    "\n",
    "class ModelTrainer:\n",
    "    def __init__(self, config: ModelTrainerConfig):\n",
    "        self.config = config\n",
    "        \n",
    "    def train(self):\n",
    "        train_data = pd.read_csv(self.config.train_data_path)\n",
    "        test_data = pd.read_csv(self.config.test_data_path)\n",
    "        \n",
    "        # split the data into x and y\n",
    "        train_x = train_data.drop(columns=[self.config.target_column])\n",
    "        test_x = test_data.drop(columns=[self.config.target_column])\n",
    "        train_y = train_data[self.config.target_column]\n",
    "        test_y = test_data[self.config.target_column]\n",
    "        \n",
    "        # train the model\n",
    "        lr = ElasticNet(alpha=self.config.alpha, l1_ratio=self.config.l1_ratio, random_state=42)\n",
    "        lr.fit(train_x, train_y)\n",
    "        \n",
    "        # save the model into the artifacts directory (model_trainer folder)\n",
    "        joblib.dump(lr, os.path.join(self.config.root_dir, f'{self.config.model_name}.joblib'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 7: update pipeline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[2024-11-20 15:05:27,648: INFO: common: yaml file: config/config.yaml loaded successfully]\n",
      "[2024-11-20 15:05:27,650: INFO: common: yaml file: params.yaml loaded successfully]\n",
      "[2024-11-20 15:05:27,652: INFO: common: yaml file: schema.yaml loaded successfully]\n",
      "[2024-11-20 15:05:27,653: INFO: common: created directory at: artifacts]\n",
      "[2024-11-20 15:05:27,654: INFO: common: created directory at: artifacts/model_trainer]\n"
     ]
    }
   ],
   "source": [
    "try:\n",
    "    config = ConfigurationManager()\n",
    "    model_trainer_config = config.get_model_trainer_config()\n",
    "    model_trainer = ModelTrainer(config=model_trainer_config)\n",
    "    model_trainer.train()\n",
    "except Exception as e:\n",
    "    raise e"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "mlproj",
   "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": "3.12.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}