Distribution Sampler

What it does

The Distribution Sampler draws reproducible random samples from all the common probability distributions — normal (Gaussian), uniform, log-normal, exponential, Poisson, binomial, Bernoulli, beta, gamma, Weibull, Pareto, chi-squared, geometric, Cauchy, Laplace, Rayleigh, logistic, hypergeometric, and more. Set the parameters, choose a sample size, optionally round or truncate the values, and reuse a seed to regenerate the exact same sample.

Common use cases

• Normal-distribution data — generate Gaussian samples with a chosen mean and standard deviation for statistics homework, chart demos, or test fixtures.
• Simulation inputs — arrival times from an exponential, event counts from a Poisson, failure times from a Weibull.
• Testing analytics code — feed your histogram, percentile, or anomaly logic a sample whose true distribution you know.
• Teaching — sample the same distribution at n=10 and n=10,000 and compare the histograms.

Settings

• Distribution — any of the supported families, from uniform and normal to heavy-tailed choices like Pareto and Cauchy.
• Parameters — mean/standard deviation, rate, shape/scale, or trial/probability values depending on the distribution; the page validates impossible combinations.
• Sample size — how many values to draw.
• Rounding — round output to a fixed number of decimals.
• Truncation range — clamp the sample to a min/max window when the tails would break your use case.
• Seed — identical seed + configuration = identical sample.

Privacy note

Samples are computed locally in your browser and never uploaded. They are synthetic numbers drawn from ideal mathematical distributions — not measurements of any real population.

FAQ

How do I generate normally distributed data? Pick “Normal”, set the mean and standard deviation, choose a sample size, and generate. About 68% of values will land within one standard deviation of the mean — visible immediately in the histogram at larger sample sizes.

Is the output reproducible? Yes. A fixed seed and configuration produce the same values every run, on any machine — ideal for golden-file tests.

Why does my small sample look “wrong”? Small samples are noisy: ten draws from a normal distribution can easily look skewed. Increase the sample size and the shape converges to the ideal curve.

What’s the difference between this and Gaussian Noise? This tool samples standalone values from a distribution you choose; the Gaussian Noise tool perturbs a signal with normal noise. For visual exploration, the Distribution Visualizer plots the curves themselves.