parameters 0.27.0

Breaking Changes

• The standardize argument in factor_analysis() now defaults to FALSE.

• The rotation argument in factor_analysis() now defaults to "oblimin",
  because the former default of "none" rarely makes sense in the context of
  factor analysis. If you want to use no rotation, please set rotation = "none".

• The cor argument in n_factors() was renamed into correlation_matrix. In
  factor_analysis(), the cor argument was completely removed to avoid naming
  collision with the cor argument of psych::fa(), which now users can pass
  the cor argument to psych::fa() when using factor_analysis().

Changes

• factor_analysis() gets a .matrix method, including arguments n_obs and
  n_matrix, to compute factor analysis for a correlation matrix or covariance
  matrix.

• New function factor_scores() to extract factor scores from EFA (psych::fa()
  or factor_analysis()).

• Added and/or improved print-methods for all functions around PCA, FA and Omega.

• Improved efficiency in model_parameters() for models from packages brms
  and rstanarm.

• p_adjust for model_parameters() gets a new options, "sup-t", to calculate
  simultaneous confidence intervals.

Bug fixes

• bootstrap_model() did not work for intercept-only models. This has been fixed.

• Fixed issue with printing labels as pretty names for models from package
  pscl, i.e. print(model_parameters(model), pretty_names = "labels") now
  works as expected.

parameters 0.26.0

Changes

• The effects argument in model_parameters() for classes merMod, glmmTMB,
  brmsfit and stanreg gets an additional "grouplevel" option, to return
  the group-level estimates for random effects.

• model_parameters() for Anova-objects gains a p_adjust argument, to apply
  p-adjustment where possible. Furthermore, for models from package afex, where
  p-adjustment was applied during model-fitting, the correct p-values are now
  returned (before, unadjusted p-values were returned in some cases).

• Revised code-base to address changes in latest insight update. Dealing with
  larger models (many parameters, many posterior samples) from packages brms
  and rstanarm is more efficient now. Furthermore, the options for the
  effects argument have a new behaviour. "all" only returns fixed effects
  and random effects variance components, but no longer the group level
  estimates. Use effects = "full" to return all parameters. This change is
  mainly to be more flexible and gain more efficiency for models with many
  parameters and / or many posterior draws.

• model_parameters() for Anova objects gains an include_intercept argument,
  to include intercepts in the Anova table, where possible.

parameters 0.25.0

Changes

• model_parameters() for objects from the marginaleffects packages now calls
  bayestestR::describe_posterior() to process Bayesian models. This offers
  more flexibility in summarizing the posterior draws from marginaleffects.

• model_parameters() now shows a more informative coefficient name for binomial
  models with probit-link.

• Argument wb_component now defaults to FALSE.

• Improved support and printing for tests from package WRS2.

Bug fixes

• Fixed printing issue with model_parameters() for htest objects when
  printing into markdown or HTML format.

• Fixed printing issue with model_parameters() for mixed models when
  include_reference = TRUE.

parameters 0.24.2

Changes

• The effects argument in model_parameters() for classes merMod, glmmTMB,
  brmsfit and stanreg gets an additional "random_total" option, to return
  the overall coefficient for random effects (sum of fixed and random effects).

Bug fixes

• Fixed issue in model_parameters() for objects from package marginaleffects
  where columns were renamed when their names equaled to certain reserved words.

parameters 0.24.1

Changes

• model_parameters() now supports objects of class survfit.

• model_parameters() now gives informative error messages for more model
  classes than before when the function fails to extract model parameters.

• Improved information for credible intervals and sampling method from output
  of model_parameters() for Bayesian models.

Bug fixes

• Fixed issue when printing model_parameters() with models from mgcv::gam().

• Fixed issues due to breaking changes in the latest release of the datawizard
  package.

• Fixed issue with wrong column-header in printed output of model_parameters()
  for MASS::polr() models with probit-link.

parameters 0.24.0

Breaking Changes

• The robust argument, which was deprecated for a long time, is now no longer
  supported. Please use vcov and vcov_args instead.

Changes

• Added support for coxph.panel models.

• Added support for anova() from models of the survey package.

• Documentation was re-organized and clarified, and the index reduced by removing
  redundant class-documentation.

Bug fixes

• Fixed bug in p_value() for objects of class averaging.

• Fixed bug when extracting 'pretty labels' for model parameters, which could
  fail when predictors were character vectors.

• Fixed bug with inaccurate standard errors for models from package fixest
  that used the sunab() function in the formula.

parameters 0.23.0

Breaking Changes

• Argument summary in model_parameters() is now deprecated. Please use
  include_info instead.

• Changed output style for the included additional information on model formula,
  sigma and R2 when printing model parameters. This information now also includes
  the RMSE.

Changes

• Used more accurate analytic approach to calculate normal distributions for
  the SGPV in equivalence_test() and used in p_significance().

• Added p_direction() methods for frequentist models. This is a convenient
  way to test the direction of the effect, which formerly was already (and still
  is) possible with pd = TRUE in model_parameters().

• p_function(), p_significance() and equivalence_test() get a vcov and
  vcov_args argument, so that results can be based on robust standard errors
  and confidence intervals.

• equivalence_test() and p_significance() work with objects returned by
  model_parameters().

• pool_parameters() now better deals with models with multiple components
  (e.g. zero-inflation or dispersion).

• Revision / enhancement of some documentation.

• Updated glmmTMB methods to work with the latest version of the package.

• Improved printing for simulate_parameters() for models from packages mclogit.

• print() for compare_parameters() now also puts factor levels into square
  brackets, like the print() method for model_parameters().

• include_reference now only adds the reference category of factors to the
  parameters table when those factors have appropriate contrasts (treatment or
  SAS contrasts).

Bug fixes

• Arguments like digits etc. were ignored in `model_parameters() for objects
  from the marginaleffects package.

parameter 0.22.2

New supported models

• Support for models glm_weightit, multinom_weightit and ordinal_weightit
  from package WeightIt.

Changes

• Added p_significance() methods for frequentist models.

• Methods for degrees_of_freedom() have been removed. degrees_of_freedom()
  now calls insight::get_df().

• model_parameters() for data frames and draws objects from package
  posterior also gets an exponentiate argument.

Bug fixes

• Fixed issue with warning for spuriously high coefficients for Stan-models
  (non-Gaussian).

parameters 0.22.1

Breaking changes

• Revised calculation of the second generation p-value (SGPV) in equivalence_test(),
  which should now be more accurate related to the proportion of the interval
  that falls inside the ROPE. Formerly, the confidence interval was simply treated
  as uniformly distributed when calculating the SGPV, now the interval is assumed
  to be normally distributed.

New supported models

• Support for svy2lme models from package svylme.

Changes

• standardize_parameters() now also prettifies labels of factors.

Bug fixes

• Fixed issue with equivalence_test() when ROPE range was not symmetrically
  centered around zero (e.g., range = c(-99, 0.1)).

• model_parameters() for anova() from mixed models now also includes the
  denominator degrees of freedom in the output (df_error).

• print(..., pretty_names = "labels") for tobit-models from package AER now
  include value labels, if available.

• Patch release, to ensure that performance runs with older version of datawizard
  on Mac OS X with R (old-release).

parameters 0.22.0

Breaking changes

• Deprecated arguments in model_parameters() for htest, aov and
  BFBayesFactor objects were removed.

• Argument effectsize_type is deprecated. Please use es_type now. This change
  was necessary to avoid conflicts with partial matching of argument names (here:
  effects).

New supported models

• Support for objects from stats::Box.test().

• Support for glmgee models from package glmtoolbox.

Bug fix

• Fixed edge case in predict() for factor_analysis().

• Fixed wrong ORCID in DESCRIPTION.