NEWS.md

In gasmodel: Generalized Autoregressive Score Models

gasmodel 0.5.1

• Added block bootstrap methods.

gasmodel 0.5.0

• Added support for summarize() and plot() generic functions.
• Added computation of p-values in the gas_bootstrap() function.
• Added functionality for parallelization in the gas_bootstrap() function.

gasmodel 0.4.0

• Added the asymmetric Laplace distribution.
• Added the Birnbaum–Saunders distribution.
• Changed the default value of par_static. Now, the first group of parameters (usually just the first parameter for univariate distributions) is dynamic while the rest are static.
• Renamed the "spec" argument to "regress". One of its value, "reg_err", renamed to "sep".
• Corrected the Fisher information for the generalized gamma and zero-inflated negative binomial distributions.
• Updated the ice_hockey_championships dataset.

gasmodel 0.3.0

• Added diagonal and full Fisher information matrix scalings.
• Added the Beta distribution.
• Added the Dirichlet distribution.
• Added the Laplace distribution.
• Added the von Mises distribution.
• Added german_car_market_cap dataset.
• Corrected NA and Inf handling for the Plackett-Luce distribution.
• Corrected the Fisher information for the generalized gamma distribution.
• Replaced the tidyverse package with dplyr, tidyr, and ggplot2 packages in Suggests.

gasmodel 0.2.0

• Added expected values, variances, and residuals to the return value of gas().
• Added the zero-inflated Skellam distribution.
• Modified the computation of the expected value, the variance, and the Fisher information of the Plackett-Luce distribution.
• Modified the computation of the Fisher information of the zero-inflated negative binomial distribution.
• Updated the sp500_daily dataset.

gasmodel 0.1.0

• Initial release on CRAN.

gasmodel documentation built on Aug. 30, 2023, 1:09 a.m.