summary.emcorrprobit: Summarizing an EMcorrProbit fit

In ninard/EMcorrProbit: Maximum likelihood estimation of correlated probit models via EM algorithm

Description

summary of an emcorrprobit object

Usage

## S3 method for class 'emcorrprobit'
summary(object, ...)
## S3 method for class 'summary.emcorrprobit'
print(x, ...)

Arguments

object	an object of class 'emcorrprobit'
x	an object of class 'summary.emcorrprobit'
...	arguments passed to or from other methods.bootstrap.samples: the number of samples used in bootsrap method, between 50 and 100 is recommended, 50 by default; doParallel: logical, if TRUE foreach function from doParallel package is used to speed up the calculation of standard errors, FALSE by default; cores: the number of cores used in parallel computation, if NULL (by default), the number of cores is set by the doParallel package; epsilon: a value for the stopping criterion. If not specified, it is defined as 10 times epsilon used for emcorrprobit fit of object.

Details

print.summary.emcorrprobit uses smart formatting of the coefficients, standard errors, etc.

Standard errors are obtained via bootstrap method and a summary table with respective z-scores and p-values is printed.

Value

The function summary.emcorrprobit computes and returns the covariance matrix of the parameters' estimates via bootstrap method. Standard errors, z-scores and p-values of the emcorrprobit estimates are presented via print function.

vcov the covariance matrix of the parameters' estimates

Examples

### using doParallel package
### for standard errors and respective P-values
#ex1.se=summary(example1, doParallel=TRUE, bootstrap.samples=50)
### print
#ex1.se
### variance-covariance matrix of the estimates
#ex1.se$vcov

### without parallel computations - very slow
#ex2.se=summary(example1, bootstrap.samples=50)
#ex2.se$vcov

ninard/EMcorrProbit documentation built on May 23, 2019, 7:05 p.m.