bizicount-package {bizicount}

bizicount: Copula-Based Bivariate Zero-Inflated Count Regression Models

Description

The package provides regression functions for copula-based bivariate count models, with and without zero-inflation, as well as regression functions for univariate zero-inflated count models. Generic methods from the texreg-package and DHARMa are extended to support this package, namely for the purposes of producing professional tables and carrying out post-estimation diagnostics. A generic for He et al. (2019)'s test for zero-modification is provided, with methods for both bizicount and glm-class objects.

Bivariate Functions

• bizicount – The primary function of this package. Carries out copula-based bivariate count regression via maximum likelihood using numerical optimization. Supports both zero-inflated and non-inflated distributions.

• extract.bizicount – Method for the texreg package's extract generic. Creates a list of texreg objects, one for each margin, for use with that package's other functions.

• make_DHARMa – Creates a list of DHARMa objects, one for each margin, for bizicount models. A wrapper around createDHARMa.

• simulate.bizicount – Method that simulates observations using the fitted model's parameters, primarily for use with DHARMa.

• zi_test – Method for testing for marginal zero-modification using the esimated parameters from the model. This test is preferable to the Vuong, Wald, Score, and LR tests. See He et al. (2019).

Univariate Functions

• zic.reg – Univariate zero-inflated count regression models via maximum likelihood.

• extract.zicreg – Method for the texreg package's extract generic. Creates a texreg object that interfaces with that package's methods.

• simulate.zicreg – Method for simulating from the fitted model. Results are generally used for creating DHARMa objects.

  #'

• zi_test – Method for testing for univariate zero-modification using the esimated parameters from the model. This test is preferable to the Vuong, Wald, Score, and LR tests. See He et al. (2019).

Author(s)

John Niehaus