Summary of a feNmlm object. Computes different types of standard errors.

Share:

Description

This function is similar to print.feNmlm. It provides the table of coefficients along with other information on the fit of the estimation. It can compute different types of standard errors. The new variance covariance matrix is an object returned.

Usage

1
2
3
## S3 method for class 'feNmlm'
summary(object,sd=c("standard","white","cluster","twoway"),
         cluster,dof_correction=TRUE,...)

Arguments

object

A feNmlm object.

sd

Character scalar. Which kind of standard error should be prompted: “standard” (default), “White”, or “cluster”?

cluster

A list of vectors. Used only if sd = "cluster" or sd="twoway". The vectors should give the cluster of each observation. Note that if the estimation was run using dummy, the standard error is automatically clustered along the cluster given in feNmlm.

dof_correction

Logical. Should a finite sample correcton be applied? (Default is TRUE.)

...

Currently unused.

Value

The same values as a feMmlm object plus:

vcov

The variance-covariance matrix whose type is the one requested by the user.

Author(s)

Laurent Berge

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
#The data
n = 100
x = rnorm(n,1,5)**2
y = rnorm(n,-1,5)**2
z = rpois(n,x*y)
base = data.frame(x,y,z)

#Comparing the results of a 'linear' function
est0L = feNmlm(z~0,base,~log(x)+log(y),family="poi")
est0NL = feNmlm(z~a*log(x)+b*log(y),base,start = list(a=0,b=0),
					family="poisson",	linear.fml=~1)

# Displaying the summary
summary(est0L,sd="white")
myWhiteVcov = summary(est0L,sd="white")$vcov