YAML
---
name: generating-test-data
description: Generates realistic test datasets and mock content using systematic approaches. Use when you need sample data for development, testing, or demonstrations.
---

Quick Start

Python
from faker import Faker
import random

fake = Faker()
users = [
    {
        "id": i,
        "name": fake.name(),
        "email": fake.email(),
        "created_at": fake.date_time_between(start_date='-2y').isoformat(),
        "status": random.choice(["active", "inactive", "pending"])
    }
    for i in range(1, 101)
]

Workflow

1. Define schema - Specify data types and relationships needed
2. Choose generation method - Faker for personal data, random for IDs/dates
3. Apply constraints - Realistic ranges, valid formats, business rules
4. Generate in batches - Create manageable chunks for large datasets
5. Export format - JSON, CSV, SQL inserts based on target system

Progress:

• Map required fields and data types
• Set up Faker with appropriate locales/providers
• Define realistic value ranges and distributions
• Generate sample batch and validate
• Scale to full dataset size

Examples

Example 1: Input: E-commerce product catalog with 500 items Output:

Python
products = [
    {
        "sku": f"PROD-{fake.random_int(10000, 99999)}",
        "name": fake.catch_phrase(),
        "price": round(random.uniform(9.99, 999.99), 2),
        "category": fake.random_element(["Electronics", "Clothing", "Books"]),
        "stock": fake.random_int(0, 100),
        "description": fake.text(max_nb_chars=200)
    }
    for _ in range(500)
]

Example 2: Input: API response simulation with nested relationships Output:

Python
{
    "user": {
        "id": 1001,
        "profile": fake.profile(),
        "orders": [
            {
                "order_id": fake.uuid4(),
                "items": fake.random_int(1, 5),
                "total": round(random.uniform(25.00, 500.00), 2),
                "status": fake.random_element(["pending", "shipped", "delivered"])
            }
            for _ in range(fake.random_int(0, 10))
        ]
    }
}

Example 3: Input: SQL test data for user authentication table Output:

SQL
INSERT INTO users (username, email, password_hash, created_at) VALUES
('john_doe_123', 'john.doe@example.com', '$2b$12$hash...', '2023-05-15 14:30:22'),
('jane_smith_456', 'j.smith@company.com', '$2b$12$hash...', '2023-06-01 09:15:11');

Best Practices

• Use Faker library for realistic personal/business data
• Set consistent seeds (Faker.seed(42)) for reproducible datasets
• Include edge cases (empty strings, null values, boundary conditions)
• Match real data distributions (80% active users, 20% inactive)
• Use appropriate locales (Faker('en_US') vs Faker('de_DE'))
• Generate related data with logical consistency (order dates after user creation)

Common Pitfalls

• Don't use sequential IDs that reveal dataset size
• Avoid unrealistic data combinations (future birth dates, negative prices)
• Don't ignore foreign key relationships in related tables
• Don't generate all data at once for large datasets (memory issues)
• Don't forget to sanitize generated data for target environment constraints