Bootstrap for Boolean Models

Share:

Description

Performs bootstrap estimates of a boolean model.

Usage

1
boolboot(obj, n = 100, method = "nlminb", cluster = NULL, ...)

Arguments

obj

boolean model object as produced by boolprep and first estimated with boolean.

n

integer specifying the number of bootstrap estimates. Defaults to 100.

method

string specifying the method of estimation. The specified method should be one of those available from the optimx or optim functions. Defaults to "nlminb".

cluster

string vector specifying hosts to use for parallel processing through parallel (see makeCluster). Defaults to NULL indicating no clustering.

...

additional parameters to pass on to subsequent functions.

Details

boolboot performs bootstrap estimated of a boolean model specified by boolprep and first estimated with boolean.

Value

boolboot returns a boolboot model object. This object is identical to a boolean model object but with an additional model.boot slot containing the results of the bootstrap. A separate object type is used to help prevent the accidential loss of bootstrap estimates.

Author(s)

Jason W. Morgan (morgan.746@osu.edu)

References

Braumoeller, Bear F. (2003) “Causal Complexity and the Study of Politics.” Political Analysis 11(3): 209–233.

See Also

See boolprep for model setup, boolean for estimation, the parallel package for details on clustering, and optimx and optim for estimation methods.