CMPParamBoot: Parametric COM-Poisson Bootstrap
Description
Generates n sample estimates for beta and nu.
Usage
CMPParamBoot(x, poissonest, betahat, nuhat, n=1000)
Arguments
x
explanatory variable data
poissonest
initial estimates from Poisson regression
betahat
COM-Poisson estimates from full data set
nuhat
COM-Poisson dispersion estimate from full data set
n
number of sample estimates to create
Value
- Ystar
- a matrix of values with n columns
- CMPresult
- a matrix of beta and nu estimates containing n rows
Details
CMPParamBoot generates n columns of simulated output
(y) data, and uses those simulations along with the explanatory (x)
data to produce n estimates for beta and nu. References
A Flexible Regression Model for Count Data, by Sellers & Shmueli, http://ssrn.com/abstract=1127359See Also
makeCMPdata, ComputeBetasAndNuHat