Most statistical procedures benefit from understanding the underlying population distribution, or at the very least offering reassurance that our assumptions about those distributions are valid. JMP provides a number of ways to easily explore and investigate distributional assumptions.
In this post, I provide information on fitting continuous or discrete distributions in the JMP Distribution platform. You can also fit, evaluate and model a wide variety of distributions in the Reliability and Survival > Life Distribution platform.
Fitting One Continuous Distribution
From an open JMP data table, select Analyze > Distribution. For this example, I use Hollywood Movies.jmp from the JMP Sample Data Directory.
Select one or more continuous variables from Select Columns, click Y, Columns, and then click OK.
Select Continuous Fit from the red triangle for the variable and select a distribution (LogNormal was selected in the example below).
In the resulting fitted distribution output, click on the red triangle and select Goodness of Fit (shown) or Diagnostic Plot to assess the fit of the distribution.
Here, the small p-value and the note provided indicate that the underlying distribution is not LogNormal.
Fitting All Continuous Distributions
To automate the process of fitting and evaluating different continuous distributions, select Continuous Fit, and then All from the red triangle for the variable. JMP will compare available continuous distributions, and will select and fit the best distribution (the distribution with the lowest AICc value). The check boxes under Compare Distributions allow you to explore the fits of the different distributions.
Fitting Discrete Distributions
If the continuous variable contains discrete values, four discrete distributions are available under Discrete Fit.
Note: For more details on fitting continuous or discrete distributions, search for Fit Distributions in the JMP Help or in the book Basic Analysis (under Help > Books).
See More in the Learning Library
Fitting Distributions is one of the many topics covered in the Learning Library. To download or view one-page guides, tutorials, short videos and other resources, visit jmp.com/learn.