Model Profiler - Profiler Tab

Select the Profiler tab of the Model Profiler to access the options described here.

Predictor Variables. This list contains all the predictor variables used to generate the models.

Distributions. This list contains appropriate random distributions available for the continuous or categorical predictor variable chosen in the Predictor Variables list.

Location. Specify the Mean parameter for the selected distribution.

Scale. Specify the Standard Deviation parameter for the selected distribution.

Min. Specify the lower bound for the chosen predictor variable. The random values simulated for the predictor variable will be greater than or equal to this value.

Max. Specify the upper bound for the chosen predictor variable. The random values simulated for the predictor variable will be less than or equal to this value.

Add Distribution. Click this button to set the distribution for the chosen predictor variable in the Predictor Variables list to the distribution chosen in the Distributions list with the specified parameter values.

Simulate correlations of inputs. Select this check box to generate a correlation matrix between the predictor variables from the input data set, and use this correlation matrix in Iman-Conover with Latin hypercube design simulation to generate the random values for the predictor variables. The Iman-Conover method will simulate the predictor variables ensuring the rank correlation structure is as close to the target correlations as possible. Clear this check box to use a Monte Carlo simulation, which assumes independence among the predictors to generate the random values for the predictor variables.

Edit Correlation. Click this button to edit the correlation matrix between predictor variables. The valid range is -1<=x<=1.

Fit Best Distributions. When you click this button, STATISTICA will fit different distributions to the continuous predictor variables to determine the best fitting distribution and set the predictor distribution parameters to the values generated from the fit.

Selected predictor distributions. Displays the distribution and the corresponding parameters used to generate values for the predictor variables.