Object Reference : Object View and Procedure Reference : Model
Stochastic solution options for models.
stochastic sets options for stochastic model solution but does not solve the model.
Note that these options have no effect on the current solve if deterministic solution has been selected.
i=arg (default=“n”)
Innovation generation: “n” (normal random number) or “b” (bootstrap).
d=arg (default=“f”)
Diagonal covariance matrix (for bootstrap: draw resids independently for each equation): “t” (true), “f” (false).
v=arg (default=“t”)
Scale covariance matrix to equation specified innovation variances: “t” (true), “f” (false). Does not apply to Bootstrap.
m=pos_number (default=1.0)
Multiply resid covariance or bootstrap by the positive number pos_number.
Covariance estimation sample (Bootstrap residual draw sample). For example, s =“1970.1 2003.4”
r=integer (default=1000)
Number of stochastic repetitions.
Calculate confidence interval from entire sample.
f=number (default=.02)
Fraction of failed repetitions before stopping.
b=number (default=.95)
Size of stochastic confidence intervals.
c=arg (default=“f”)
Include coefficient uncertainty: “t” (true), “f” (false).
Page name for a new workfile page to save the results of all repetitions of the stochastic solve. If blank (default) only summaries (mean, sd, etc.) of the repetitions are maintained.
See “Models” for a discussion of models. See Model::innov to set options on individual series in stochastic solution.
See also Model::model, Model::solve and Model::solveopt.