Spotfire Statistica® Distributions & Simulation

The purpose of the Distributions & Simulation analysis is to provide a general tool for performing simulation studies. Specifically, this module will enable users to perform "modern design of experiments" (see Giunta, Wojtkiewicz, Eldred 2003) by simulating multivariate design (or input) variables from specific distributions, and (rank-order) covariances that define the space of interest. Although seemingly simple, this module enables you to accurately model the current processes that generate the data, and from there you can simulate those processes, and evaluate the performance of a system. 

The following distributions are available; Normal, Log-Normal, Folded normal, half normal, Rayleigh, Weibull, Gaussian mixture, Johnson distribution (general non-normal, single mode), Generalized Extreme Value, Generalized Pareto, Triangular, Binomial, Poisson, Geometric, Bernoulli, Histogram. 

Distribution summary statistics include the Kolmogorov-Smirnov (KS) statistic and the Anderson-Darling statistic to evaluate the fit of different distributions.  The KS statistic and p-values reported are based on those tabulated by Massey (1951). The critical values for the Anderson-Darling statistic have been tabulated (see, for example, Dodson, 1994, Table 4.4) for sample sizes between 10 and 40; however, the critical values (and p-values) reported in Distributions & Simulations are calculated via an approximation method (Marsaglia, 2004).

Data can be simulated via Monte Carlo, Latin Hypercube Sampling (LHS), Iman Conover, and LHS with Iman Conover.

The model profiler feature runs simulations based on a specified model in the Design of Experiments (DOE), General Linear Models (GLM), or Distributions & Simulation modules.