Paste a set of data; the tool compares fits across 6 distributions using the Anderson-Darling statistic (smaller = better fit) to help you decide which distribution to use for non-normal Cpk / reliability. Weibull / lognormal / exponential / Gamma are included only when all data are positive.