Fit/Setup

The Fit Setup dialog provides a list of all distributions supported by Stat::Fit and the relevant choices for goodness of fit tests.   At least one distribution must be chosen before the estimate, test, and graphing commands become available.   

Note the the Fit Setup dialog is available both before and after a document is opened.   If no document is open, the Fit Setup dialog applies to the overall program.   A new document assumes the values stored in this main Fit Setup dialog, and, thereafter, can be customized.

At any time the default Fit Setup dialog can be changed by checking the Save To Default check box at the bottom of the dialog and clicking on OK.   This default configuration is loaded in the Fit Setup dialog when the program starts.

Stat::Fit does not start the fitting process when this dialog is exited, but turns on the appropriate commands and waits for further instruction.

        distribution choice

The Distribution page of the Fit Setup dialog provides a distribution list for the choice of distributions for subsequent fitting.   All distributions chosen here will be used sequentially for estimates and goodness of fit tests.   Clicking on a distribution name in the distribution list on the left chooses that distribution and moves that distribution name to the distributions selected box on the right unless it is already there.   Clicking on the distribution name in the distributions selected box on the right removes the distribution.   All distributions may be moved to the distributions selected box by clicking the Select All button.   The distribution selected box may be cleared by clicking the Clear button

        goodness of fit choice

The Calculations page of the Fit Setup dialog allows the appropriate goodness of fit tests to be chosen, along with supporting parameters.

All goodness of fit tests use either the moments estimates or the maximum likelihood estimates for each selected distribution as set here.   This does not prevent either set of estimates from being calculated, but designates which estimates will be used for the goodness of fit tests and result graphs.   The default is the maximum likelihood estimates.

For continuous distributions with a lower bound or minimum such as the Exponential, the lower bound can be forced to assume a value at or below the minimum data value.   This lower bound will be used for both the moments and maximum likelihood estimates.   By default, it is pinned at the lowest integer below the minimum data point.   If new data is added below a preset lower bound, the bound will be modified to assume the closest integer value below all input data.

The accuracy of the estimation is set, by default, to 0.0003, but can be changed if greater accuracy is desired.   Note that greater accuracy can mean much greater calculation time.   Some distributions have either moments estimates and/or maximum likelihood estimates which do not require iterative estimation; in these cases, the accuracy will not make any difference in the estimation.

The Level of Significance is used for all goodness of fit tests, and is set, by default, to 0.05.

The goodness of fit tests include the Chi Squared test, the Kolmogorov Smirnov test, and the Anderson Darling test.   The test selection is set, by default, to the Chi Squared test.   The Chi Squared test may be done with intervals of equal length or equal probability.   The interval type is set to equal probability, by default, because this test gives more consistent results with peaked distributions.