The COMPoissonReg R package fits COM-Poisson regression (Sellers & Shmueli, 2010) and zero-inflated (ZI) COM-Poisson regression models (Sellers & Raim, 2016). The syntax to fit a model looks like
cmp.out <- glm.cmp(formula.lambda = y ~ x1 + x2, formula.nu = ~ x1, formula.p = ~ x2)where formula.lambda is a regression model for the COM-Poisson rate parameter, formula.nu is a model for the dispersion parameter, and formula.p is a model for the zero-inflation parameter. See the package documentation for details.
- Kimberly F. Sellers and Galit Shmueli (2010). A Flexible Regression Model for Count Data. Annals of Applied Statistics, 4(2), 943-961. [http]
- Kimberly F. Sellers and Andrew M. Raim (2016). A Flexible Zero-Inflated Model to Address Data Dispersion. Computational Statistics and Data Analysis, 99, 68-80. [http]