The HPCOUNTREG procedure estimates the parameters of a count regression model by maximum likelihood techniques.
The HPCOUNTREG procedure supports the following models for count data:
Poisson regression
Conway-Maxwell-Poisson regression
negative binomial regression with quadratic and linear variance functions (Cameron and Trivedi 1986)
zero-inflated Poisson (ZIP) model (Lambert 1992)
zero-inflated Conway-Maxwell-Poisson (ZICMP) model
zero-inflated negative binomial (ZINB) model
fixed-effects and random-effects Poisson models for panel data
fixed-effects and random-effects negative binomial models for panel data
The following list summarizes some basic features of the HPCOUNTREG procedure:
can perform analysis on a massively parallel high-performance appliance
reads input data in parallel and writes output data in parallel when the data source is the appliance database
is highly multithreaded during all phases of analytic execution
has model-building syntax that uses CLASS and effect-based MODEL statements familiar from SAS/ETS analytic procedures
performs maximum likelihood estimation
supports multiple link functions
uses the WEIGHT statement for weighted analysis
uses the FREQ statement for grouped analysis
uses the OUTPUT statement to produce a data set that contains predicted probabilities and other observationwise statistics