Learn R Programming
COMPoissonReg (version 0.3.5)

cmp: Functions to estimate COM-Poisson model parameters (betas and nu)

Description

Estimates maximum likelihood estimates for betas and nu using nlminb.

Usage

cmp(formula, initial.est=NULL, nuinit=1, max=100, ...)


	## S3 function for class 'cmp'
	"nu"(object, ...)


	## S3 function for class 'cmp'
	"chisq"(object, ...)


	## S3 function for class 'cmp'
	"pval"(object, ...)


	## S3 function for class 'cmp'
	"sdev"(object, ...)


	## S3 function for class 'cmp'
	"deviance"(object, ...)


	## S3 function for class 'cmp'
	"leverage"(object, ...)


	## S3 function for class 'cmp'
	"predict"(object, ...)


	## S3 function for class 'cmp'
	"residuals"(object, ...)


	## S3 function for class 'cmp'
	"parametric_bootstrap"(object, ...)

Arguments

formula
formula for the COM-Poisson model
initial.est
initial vector of betas, b0_1, ..., b0_p; if NULL, estimated using Poisson GLM
nuinit
initial value for nu
max
maximum number to use for truncating infinite sums
...
other model parameters, such as data
object
object of type 'cmp'

Value

An object of class "cmp", from which the coefficients and other information can be computed.

Details

cmp finds the maximum likelihood estimates for the COM-Poisson estimators (betas and nu) using the function ComputeBetasAndNuHat.

References

A Flexible Regression Model for Count Data, by Sellers & Shmueli, http://ssrn.com/abstract=1127359

See Also

COMPoissonReg ComputeBetasAndNuHat