Skip to contents

Changelog

Source: NEWS.md

TDMM 0.3.0

  • tdmm() and tdmm.parallel() support models with or without a subject-specific random intercept through random_intercept; TRUE remains the default.
  • Gaussian, Bernoulli, and Poisson simulation helpers generate independent-observation data directly when random_intercept = FALSE.
  • Repeated-observation models with a subject-specific random intercept remain supported.

TDMM 0.2.0

  • Fit Gaussian, Bernoulli, and Poisson longitudinal outcomes with subject-specific irregular observation times, unequal visit counts, baseline covariates, and time-varying covariates.
  • Use cubic B-splines (degree 3, order 4) with a second-order coefficient-difference penalty and eight basis functions by default. n_basis specifies the basis dimension directly; the older nknots argument remains supported with a deprecation warning.
  • Fit serial or parallel chains through tdmm() and structured tdmm_control() and tdmm.priors() settings. tdmm.parallel() remains a compatibility wrapper.
  • Existing Gaussian and Poisson model specifications are retained.

TDMM 0.1.0

  • Initial archived release. See the v0.1.0 release for its source and documentation.