Probability Distribution Matcher

Identify which probability distribution best fits your data with goodness-of-fit tests

Overview

The Probability Distribution Matcher helps you identify which probability distribution best describes your data through visual comparison and statistical testing. Upload your own data or use synthetic datasets to compare against common distributions like Normal, Binomial, Poisson, Exponential, and many others using goodness-of-fit tests.

Tips

1. Start with visualization: Before running statistical tests, examine histograms and basic statistics to narrow down candidate distributions based on shape and domain.

2. Consider data context: Let the data-generating process guide your choice - count data suggests Poisson or Binomial, waiting times suggest Exponential, and symmetric continuous data suggests Normal.

3. Use Q-Q plots for assessment: Quantile-quantile plots provide intuitive visual confirmation of fit quality and reveal exactly where deviations occur.

4. Test multiple candidates: Don’t stop at the first distribution with a good p-value - compare several plausible options and prefer simpler models when fits are similar.

5. Validate with out-of-sample data: If possible, test your chosen distribution on new data to confirm it wasn’t overfitted to your initial sample.