In One Dimension
SAS/INSIGHT software enables you to explore distributions quickly. Just point and click to compute descriptive statistics or to create interactive histograms, box plots, and quantile-quantile plots.
Box Plot with Descriptive Statistics
Quantile-Quantile Plot
You can fit parametric density estimates and create quantile-quantile plots for normal, lognormal, exponential, and Weibull distributions. Add kernel density estimates based on normal, triangular, and quadratic kernels. Use slider controls to adjust curve fits dynamically.
Parametric and Kernel Density Estimation
Use Kolmogorov's D statistic to test for normal, lognormal, exponential, and Weibull distributions. For completely specified distributions, you can plot the cumulative distribution function with confidence bands to show test results graphically as well as numerically.
Testing for Normal Distribution
For robust estimation, calculate trimmed means, Winsorized means, and Gini's mean difference. To compare distributions, create side-by-side box plots.
Comparing Distributions
Statistics and Operations Research Home Page | SAS/INSIGHT Software