One powerful yet often overlooked application of generative AI is creating test data samples. Instead of manually crafting data sets, you can leverage tools like ChatGPT to generate realistic sample data quickly and efficiently.

This approach is especially useful when you’re learning frameworks like Python’s pandas and need a structured dataset to experiment with. By using a well-crafted prompt (or even not well crafted), you can produce diverse data samples tailored to your learning objectives.

For example, here’s a prompt I used with ChatGPT 4o to generate sample data:

I want to generate code for a test python DataFrame. The table has two columns Flavor and Scoops. The flavor column is persons favorite flavor. Number of scoops should stay between 1 and 3 scoops. Create 5 common flavors to be used for the test sample. Create 30 examples in the dataframe.

And this was the result.

import pandas as pd
import random

# Define common flavors and scoop range
flavors = ["Vanilla", "Chocolate", "Strawberry", "Mint Chocolate Chip", "Cookie Dough"]

# Generate sample data
data = {
    "Flavor": [random.choice(flavors) for _ in range(30)],
    "Scoops": [random.randint(1, 3) for _ in range(30)]
}

# Create DataFrame
df = pd.DataFrame(data)