Python Tutorial: Generate Random Numbers and Data Using the random Module

Python’s built-in `random` module makes it easy to generate realistic-looking dummy data—random numbers, random selections from lists, weighted outcomes, shuffling, and unique samples—without installing any third-party packages. The key takeaway is that `random` is ideal for games, simulations, and test data, but it’s not meant for security or cryptography; for that, Python recommends using the `secrets` module instead.

The tutorial starts with importing `random` from the standard library. It then walks through the most common number-generation methods. `random()` returns a float between 0 and 1, where 0 is inclusive and 1 is non-inclusive—so exact 1 never appears. For bounded floats, `uniform(a, b)` produces a random floating-point value between two endpoints (by changing the arguments, any range can be used). For integers, `randint(a, b)` simulates discrete outcomes such as a six-sided die, including both endpoints; the same method can model a coin toss by treating 0 as heads and 1 as tails.

Selection from collections is handled with `choice()` and `choices()`. `choice(sequence)` picks a single random element from a list. `choices(sequence, k=...)` returns multiple picks and can repeat elements, which is useful for simulating repeated events like roulette spins. The tutorial then adds realism with weighted probabilities: by supplying a `weights` list, outcomes can match a non-uniform distribution. Using roulette as the example, red and black are weighted 18 each while green is weighted 2, reflecting the typical pocket counts; the resulting selections show green appearing far less often.

For rearranging data, `shuffle()` randomizes a list in place—illustrated by treating numbers 1 through 52 as a deck of cards and shuffling them. To draw unique cards, `sample(population, k)` is used instead of `choices()`, because `sample()` guarantees no repeats within the draw. The example draws 5 distinct values from the 52-card deck, mirroring a hand of cards.

Finally, the tutorial connects these tools to a practical workflow: generating fake datasets for CSV-style practice. It builds lists of first names, last names, street names, and fake cities/states, then loops to create many records. Phone numbers and addresses are assembled with f-strings using random integer ranges for numeric parts and `choice()` for street/city/state parts. Emails are generated by combining a random first name with a lowercased last name and appending a fixed bogus domain. In roughly a few dozen lines, the approach produces dozens of plausible names, phone numbers, addresses, and emails—useful for testing code without needing real personal data.