Interactive Lesson: Python Food Delivery

Markdown

🍕 Welcome to Food Delivery Analytics!

In this interactive notebook, you’ll analyze real-world food delivery data to uncover insights about restaurant performance, delivery times, and customer satisfaction.

📊 Dataset Overview:
• 10,000+ food delivery orders
• Restaurant ratings and cuisines
• Delivery times and distances
• Customer feedback and tips

Learning Objectives:

Master Python data structures (lists, dictionaries, tuples)
Use Pandas for data manipulation and analysis
Clean and transform messy data
Perform grouping and aggregations
Build a complete ETL pipeline

Code In [1]:

# Exercise 1: Python Basics – Working with Restaurant Data
# Let’s start by creating basic data structures for our food delivery system

# Create a list of popular cuisines
cuisines = [“Italian”, “Chinese”, “Mexican”, “Indian”, “Thai”]

# Create a dictionary for a restaurant
restaurant = {
    “name”: “Pizza Paradise”,
    “cuisine”: “Italian”,
    “rating”: 4.5,
    “delivery_time”: 30,
    “is_open”: True
}

# Use f-strings to create a description
description = f”{restaurant[‘name’]} serves {restaurant[‘cuisine’]} food with a {restaurant[‘rating’]}★ rating”
print(description)

# Create a tuple for coordinates (immutable)
restaurant_location = (37.7749, -122.4194)  # San Francisco
print(f”Location: {restaurant_location[0]}°N, {restaurant_location[1]}°W”)
                

Code In [2]: EXERCISE

Hints:
• Use cuisines.append("Japanese") to add items
• Dictionary format: {"key": "value", "key2": value2}
• Sets use curly braces: {"Zone1", "Zone2", "Zone3"}

Code In [3]:

# Exercise 2: Introduction to Pandas – Loading Delivery Data
import pandas as pd
import numpy as np

# Create sample delivery data
delivery_data = {
    ‘order_id’: [1001, 1002, 1003, 1004, 1005],
    ‘restaurant’: [‘Pizza Paradise’, ‘Sushi Supreme’, ‘Taco Town’, ‘Burger Barn’, ‘Pizza Paradise’],
    ‘cuisine’: [‘Italian’, ‘Japanese’, ‘Mexican’, ‘American’, ‘Italian’],
    ‘order_value’: [25.99, 45.50, 18.75, 22.00, 31.25],
    ‘delivery_time’: [28, 35, 22, 25, 32],
    ‘rating’: [5, 4, 5, 3, 4],
    ‘tip’: [5.00, 8.00, 3.00, 2.00, 6.00]
}

# Create DataFrame
df = pd.DataFrame(delivery_data)

# Display basic information
print(“🍔 Food Delivery Orders Dataset”)
print(“=” * 40)
print(df.head())
print(f”\nDataset shape: {df.shape}“)
print(f”Total orders: {len(df)}“)
                

Code In [4]: EXERCISE

Code In [5]:

# Exercise 4: Data Filtering & Selection

# Expand our dataset
np.random.seed(42)
expanded_data = {
    ‘order_id’: range(2001, 2021),
    ‘restaurant’: np.random.choice([‘Pizza Paradise’, ‘Sushi Supreme’, ‘Taco Town’, 
                                     ‘Burger Barn’, ‘Thai Terrace’], 20),
    ‘order_value’: np.random.uniform(15, 60, 20).round(2),
    ‘delivery_time’: np.random.randint(15, 45, 20),
    ‘rating’: np.random.choice([3, 4, 5], 20, p=[0.2, 0.3, 0.5]),
    ‘is_peak_hour’: np.random.choice([True, False], 20)
}

df_orders = pd.DataFrame(expanded_data)

# Filter high-value orders
high_value = df_orders[df_orders[‘order_value’] > 40]
print(f”High-value orders (>$40): {len(high_value)}“)
print(high_value[[‘restaurant’, ‘order_value’]])

# Filter by multiple conditions
fast_good = df_orders[
    (df_orders[‘delivery_time’] < 25) & 
    (df_orders[‘rating’] >= 4)
]
print(f”\nFast & highly-rated deliveries: {len(fast_good)}“)
                

Code In [6]: CHALLENGE

                    </div>
        </div>

        <!-- Exercise 6: Grouping and Aggregation -->
        <div class="cell">
            <div class="cell-header">
                <span class="cell-type code">Code</span>
                <span class="cell-number">In [7]:</span>
                <button class="run-button" onclick="pyRunCell(7)">▶ Run</button>
            </div>
            <div class="code-content">
                <div class="code-editor">
<span class="comment"># Exercise 6: Grouping & Aggregations – Restaurant Performance Analysis</span>

<span class="comment"># Create comprehensive dataset</span>
<span class="keyword">np.random.seed</span>(<span class="number">42</span>)
<span class="keyword">n_orders</span> = <span class="number">100</span>

<span class="keyword">restaurants</span> = [<span class="string">‘Pizza Paradise’</span>, <span class="string">‘Sushi Supreme’</span>, <span class="string">‘Taco Town’</span>, <span class="string">‘Burger Barn’</span>, <span class="string">‘Thai Terrace’</span>]
<span class="keyword">full_data</span> = {
    <span class="string">‘order_id’</span>: <span class="keyword">range</span>(<span class="number">5001</span>, <span class="number">5001</span> + n_orders),
    <span class="string">‘restaurant’</span>: np.random.choice(restaurants, n_orders),
    <span class="string">‘order_value’</span>: np.random.uniform(<span class="number">10</span>, <span class="number">75</span>, n_orders).round(<span class="number">2</span>),
    <span class="string">‘delivery_time’</span>: np.random.normal(<span class="number">30</span>, <span class="number">8</span>, n_orders).round().astype(int),
    <span class="string">‘rating’</span>: np.random.choice([<span class="number">1</span>, <span class="number">2</span>, <span class="number">3</span>, <span class="number">4</span>, <span class="number">5</span>], n_orders, p=[<span class="number">0.05</span>, <span class="number">0.1</span>, <span class="number">0.2</span>, <span class="number">0.35</span>, <span class="number">0.3</span>]),
    <span class="string">‘day_of_week’</span>: np.random.choice([<span class="string">‘Mon’</span>, <span class="string">‘Tue’</span>, <span class="string">‘Wed’</span>, <span class="string">‘Thu’</span>, <span class="string">‘Fri’</span>, <span class="string">‘Sat’</span>, <span class="string">‘Sun’</span>], n_orders)
}

<span class="dataframe">df_full</span> = pd.DataFrame(full_data)

<span class="comment"># Group by restaurant and analyze</span>
<span class="dataframe">restaurant_stats</span> = <span class="dataframe">df_full</span>.groupby(<span class="string">‘restaurant’</span>).agg({
    <span class="string">‘order_value’</span>: [<span class="string">‘mean’</span>, <span class="string">‘sum’</span>, <span class="string">‘count’</span>],
    <span class="string">‘delivery_time’</span>: <span class="string">‘mean’</span>,
    <span class="string">‘rating’</span>: <span class="string">‘mean’</span>
}).round(<span class="number">2</span>)

<span class="function">print</span>(<span class="string">“📊 Restaurant Performance Dashboard”</span>)
<span class="function">print</span>(<span class="string">“=”</span> * <span class="number">50</span>)
<span class="function">print</span>(<span class="dataframe">restaurant_stats</span>)

<span class="comment"># Find best performer</span>
<span class="keyword">best_restaurant</span> = <span class="dataframe">restaurant_stats</span>[(<span class="string">‘rating’</span>, <span class="string">‘mean’</span>)].idxmax()
<span class="function">print</span>(<span class="string">f”\n🏆 Best rated restaurant: </span><span class="keyword">{best_restaurant}</span><span class="string">“</span>)
                </div>
            </div>
            <div class="output" id="output7"></div>
        </div>

        <!-- Exercise 7: Building an ETL Pipeline -->
        <div class="cell">
            <div class="cell-header">
                <span class="cell-type code">Code</span>
                <span class="cell-number">In [8]:</span>
                <span class="exercise-badge">FINAL PROJECT</span>
                <button class="run-button" onclick="pyRunCell(8)">▶ Run</button>
            </div>
            <div class="code-content">
                <div class="code-editor editable">
                    <textarea class="code-input" id="code8" placeholder="# FINAL PROJECT: Build a Complete ETL Pipeline
# Extract, Transform, and Load delivery data for analytics

def extract_data():
    '''Extract: Simulate loading data from multiple sources'''
    # Source 1: Orders
    orders = pd.DataFrame({
        'order_id': range(6001, 6011),
        'customer_id': np.random.randint(1, 6, 10),
        'restaurant_id': np.random.randint(1, 4, 10),
        'amount': np.random.uniform(15, 60, 10).round(2),
        'timestamp': pd.date_range('2024-01-01', periods=10, freq='H')
    })
    
    # Source 2: Restaurants
    restaurants = pd.DataFrame({
        'restaurant_id': [1, 2, 3],
        'name': ['Pizza Palace', 'Burger Barn', 'Sushi Supreme'],
        'cuisine': ['Italian', 'American', 'Japanese'],
        'avg_prep_time': [20, 15, 25]
    })
    
    return orders, restaurants

def transform_data(orders, restaurants):
    '''Transform: Clean and enrich the data'''
    # YOUR TASK: Complete the transformation
    
    # 1. Merge orders with restaurant data
    merged = orders.merge(_____, on=_____, how=_____)
    
    # 2. Add calculated fields
    merged['hour_of_day'] = merged['timestamp'].dt._____
    merged['is_dinner'] = merged['hour_of_day'] >= _____  # After 5 PM
    
    # 3. Categorize order sizes
    merged[‘order_size’] = pd.cut(
        merged[‘amount’],
        bins=[0, 20, 40, 100],
        labels=[‘Small’, ‘Medium’, ‘Large’]
    )
    
    # 4. Calculate metrics
    summary = merged.groupby(‘cuisine’).agg({
        ‘amount’: [‘mean’, ‘sum’],
        ‘order_id’: ‘count’
    })
    
    return merged, summary

def load_data(transformed_data, summary):
    ”’Load: Output results (simulate database write)”’
    print(‘📦 ETL Pipeline Results’)
    print(‘=’ * 50)
    print(‘Transformed Data Sample:’)
    print(transformed_data.head())
    print(‘\n📊 Summary by Cuisine:’)
    print(summary)
    
    # YOUR TASK: Add success metrics
    total_orders = len(_____)
    total_revenue = transformed_data[‘_____’].sum()
    print(f’\n✅ ETL Complete: {total_orders} orders, ${total_revenue:.2f} total’)
    
    return True

# Run the pipeline
orders, restaurants = extract_data()
transformed, summary = transform_data(orders, restaurants)
success = load_data(transformed, summary)”>
                

Markdown

🎯 Knowledge Check Quiz

Test your understanding of Python and Pandas concepts!

Question 1: Which data structure would be best for storing unique customer IDs?

List

Set

Tuple

Dictionary

Question 2: What Pandas method would you use to find missing values?

.dropna()

.isna()

.fillna()

.notna()

Markdown

🎉 Congratulations!

You’ve completed the Python & Pandas Food Delivery Analytics exercise!

Key Skills Practiced:
✅ Python data structures (lists, dicts, tuples, sets)
✅ Pandas DataFrames and Series
✅ Data cleaning and transformation
✅ Grouping and aggregations
✅ Building ETL pipelines
✅ Best practices for analytics engineering

Next Steps:

Practice with larger datasets
Explore advanced Pandas functions
Learn about data visualization with matplotlib/seaborn
Build automated reporting pipelines
Integrate with SQL databases