Fit Details (simulation)

The Fit Details dialog displays the results of automatic distribution fitting for a particular input. Distributions are ordered by goodness of fit, with the closest fitting distribution listed first. You can override the closest fitting distribution by selecting the radio button for the distribution you want in the Use column. Selecting a radio button in the Use column also displays a plot of the distribution superimposed on a histogram (or bar chart) of the historical data for that input.

Fit statistics. By default and for continuous fields, the Anderson-Darling test is used for determining goodness of fit. Alternatively, and for continuous fields only, you can specify the Kolmogorov-Smirnoff test for goodness of fit by selecting that choice on the Advanced Options settings. For continuous inputs, results of both tests are shown in the Fit Statistics column (A for Anderson-Darling and K for Kolmogorov-Smirnoff), with the chosen test used to order the distributions. For ordinal and nominal inputs the chi-square test is used. The p-values associated with the tests are also shown.

Parameters. The distribution parameters associated with each fitted distribution are displayed in the Parameters column. Parameters for the following distributions have the same meaning as in the associated random variable functions available in the Compute Variable dialog box: Bernoulli, beta, binomial, exponential, gamma, lognormal, negative binomial (trials and failures), normal, Poisson and uniform. See the topic Random variable functions for more information. For the categorical distribution, the parameter names are the categories and the parameter values are the associated probabilities.

Refitting with a customized distribution set. By default, the measurement level of the input is used to determine the set of distributions considered for automatic distribution fitting. For example, continuous distributions such as lognormal and gamma are considered when fitting a continuous input but discrete distributions such as Poisson and binomial are not. You can choose a subset of the default distributions by selecting the distributions in the Refit column. You can also override the default set of distributions by selecting a different measurement level from the Treat as (Measure) dropdown list and selecting the distributions in the Refit column. Click Run Refit to refit with the custom distribution set.

Note:
  • Cases with missing values for any simulated input are excluded from distribution fitting, computation of correlations, and computation of the optional contingency table (for inputs with a Categorical distribution). You can optionally specify whether user-missing values of inputs with a Categorical distribution are treated as valid. By default, they are treated as missing. For more information, see the topic Advanced Options (simulation).
  • For continuous and ordinal inputs, if an acceptable fit cannot be found for any of the tested distributions, then the Empirical distribution is suggested as the closest fit. For continuous inputs, the Empirical distribution is the cumulative distribution function of the historical data. For ordinal inputs, the Empirical distribution is the categorical distribution of the historical data.

To access the Fit Details dialog

  1. From the Simulated Fields settings on the Simulation tab, click Fit All or select the input and click Fit.
  2. Select the input and click Fit Details.