{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Question 1 SOLUTION: Predict Food Delivery Time\n",
    "\n",
    "You are provided with a food delivery dataset that includes features such as delivery distance, restaurant preparation time, traffic conditions, and order details. Your objective is to develop a regression model that accurately predicts the delivery time in minutes.\n",
    "\n",
    "Your target is the column: \"delivery_time\".\n",
    "\n",
    "You are provided with the code to download and load the csv file.\n",
    "\n",
    "Your work will be evaluated based on the completion of the following tasks:\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 1: Read Data\n",
    "\n",
    "**Tasks:**\n",
    "\n",
    "1. Read the dataset using `read_csv()`\n",
    "2. Inspect the first few rows using `head()`\n",
    "3. Display dataset information using `info()`\n",
    "4. Show statistical description using `describe()`\n",
    "5. Plot the target distribution (delivery_time)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import kagglehub\n",
    "\n",
    "# Download latest version\n",
    "path = kagglehub.dataset_download(\"mohammad2012191/q1-ka-ai-2026\")\n",
    "\n",
    "print(\"Path to dataset files:\", path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import os\n",
    "df = pd.read_csv(os.path.join(path, 'Q1_data.csv'))\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['Delivery_Time'].hist(bins=30)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 2: Data Cleaning\n",
    "\n",
    "**Tasks:**\n",
    "\n",
    "Inspect and fix the following when needed:\n",
    "\n",
    "1. **Drop the 'Order_ID' column from the data** \n",
    "2. **Handle missing values appropriately** (Hint: I guess you want to have a closer look at the columns with missing values :) )\n",
    "3. **Check and remove duplicates** if any exist\n",
    "4. **Encode categorical variables** if needed\n",
    "5. **Apply feature scaling** to numerical features (Use StandardScaler)\n",
    "6. **Check for target imbalance and state if it is imbalanced or not** (keep this cell empty if not needed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Task 1\n",
    "df = df.drop('Order_ID', axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Task 2\n",
    "# Drop rows with missing target\n",
    "df = df.dropna(subset=['Delivery_Time'])\n",
    "# Fill missing values\n",
    "df = df.fillna(df.median(numeric_only=True))\n",
    "df = df.fillna(df.mode().iloc[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Task 3\n",
    "df = df.drop_duplicates()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Task 4\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "categorical_cols = df.select_dtypes(\"object\").columns\n",
    "for col in categorical_cols:\n",
    "    le = LabelEncoder()\n",
    "    df[col] = le.fit_transform(df[col])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Task 5\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "scaler = StandardScaler()\n",
    "X_cols = [col for col in df.columns if col != 'Delivery_Time']\n",
    "df[X_cols] = scaler.fit_transform(df[X_cols])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Task 6\n",
    "# Not applicable for regression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 3: Modeling\n",
    "\n",
    "**Tasks:**\n",
    "\n",
    "1. Split the dataset into features (X) and target (y)\n",
    "2. Use the correct split: **KFold** OR **StratifiedKFold**\n",
    "3. Train a **RandomForest** model\n",
    "4. Evaluate using **MAE (Mean Absolute Error) ONLY**\n",
    "5. Print the averaged score across all folds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df.drop('Delivery_Time', axis=1)\n",
    "y = df['Delivery_Time']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import KFold\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "import numpy as np\n",
    "\n",
    "kf = KFold(n_splits=5, shuffle=True, random_state=42)\n",
    "model = RandomForestRegressor(random_state=42)\n",
    "scores = []\n",
    "\n",
    "for train_idx, val_idx in kf.split(X):\n",
    "    X_train, X_val = X.iloc[train_idx], X.iloc[val_idx]\n",
    "    y_train, y_val = y.iloc[train_idx], y.iloc[val_idx]\n",
    "    model.fit(X_train, y_train)\n",
    "    y_pred = model.predict(X_val)\n",
    "    scores.append(mean_absolute_error(y_val, y_pred))\n",
    "\n",
    "print(f'Average MAE: {np.mean(scores):.2f}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 4: Plots\n",
    "\n",
    "**Tasks:**\n",
    "\n",
    "1. Plot feature importance from your trained model\n",
    "2. Plot predicted delivery time histogram"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "plt.barh(X.columns, model.feature_importances_)\n",
    "plt.xlabel('Importance')\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = model.predict(X)\n",
    "plt.hist(y_pred, bins=30)\n",
    "plt.xlabel('Predicted Delivery Time')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 5: Bonus - Ensemble\n",
    "\n",
    "**Task:**\n",
    "\n",
    "Why use one model? Let's use 2 then merge!\n",
    "\n",
    "Rewrite the KFold loop such that you:\n",
    "1. Train **2 different models** inside the loop at the same time in each iteration\n",
    "2. Average their predictions\n",
    "3. Calculate the MAE metric on the averaged predictions\n",
    "\n",
    "Use RandomForestRegressor and CatBoostRegressor."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from catboost import CatBoostRegressor\n",
    "\n",
    "kf = KFold(n_splits=5, shuffle=True, random_state=42)\n",
    "model1 = RandomForestRegressor(random_state=42)\n",
    "model2 = CatBoostRegressor(verbose=0, random_state=42)\n",
    "scores = []\n",
    "\n",
    "for train_idx, val_idx in kf.split(X):\n",
    "    X_train, X_val = X.iloc[train_idx], X.iloc[val_idx]\n",
    "    y_train, y_val = y.iloc[train_idx], y.iloc[val_idx]\n",
    "    \n",
    "    model1.fit(X_train, y_train)\n",
    "    model2.fit(X_train, y_train)\n",
    "    \n",
    "    y_pred1 = model1.predict(X_val)\n",
    "    y_pred2 = model2.predict(X_val)\n",
    "    y_pred_avg = (y_pred1 + y_pred2) / 2\n",
    "    \n",
    "    scores.append(mean_absolute_error(y_val, y_pred_avg))\n",
    "\n",
    "print(f'Average MAE (Ensemble): {np.mean(scores):.2f}')"
   ]
  }
 ],
 "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": "3.9.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}