ComputeLambdaAndNuHat.shiftCMPest:
Compute Lambda and Nu Hat for a given data set
Description
Computes the COM-Poisson lambda and nu parameter estimates for a given data set.
Usage
ComputeLambdaAndNuHat.shiftCMPest(y,lambdainit=mean(y)-min(y),nuinit=1,a=min(y),max=100)
Arguments
y
The data set in question, as a vector
lambdainit
Initial value for Lambda
nuinit
Initial value for Nu
a
The location shift constant
max
Maximum number of iterations to run for truncating infinite sums
Value
Returns the maximum likelihood estimates for lambda and nu using constrained optimization through nlminb.
Details
Uses the built-in nlminb function (constrained optimization) to maximize the COM-Poisson log-likelihood function for a given data set and therefore compute the maximum likelihood estimators Lambda and Nu for said data set.
References
Sellers, Kimberly F. "A Generalized Statistical Control Chart for Over- or Under-Dispersed Data." Quality and Reliability Engineering International 2012. ; 28:59-65.