Changelog
Source:NEWS.md
parameters 0.24.0
parameters 0.23.0
CRAN release: 20241018
Breaking Changes
Argument
summary
inmodel_parameters()
is now deprecated. Please useinclude_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 inp_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 withpd = TRUE
inmodel_parameters()
.p_function()
,p_significance()
andequivalence_test()
get avcov
andvcov_args
argument, so that results can be based on robust standard errors and confidence intervals.equivalence_test()
andp_significance()
work with objects returned bymodel_parameters()
.pool_parameters()
now better deals with models with multiple components (e.g. zeroinflation 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()
forcompare_parameters()
now also puts factor levels into square brackets, like theprint()
method formodel_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).
parameters 0.22.2
CRAN release: 20240903
New supported models
 Support for models
glm_weightit
,multinom_weightit
andordinal_weightit
from package WeightIt.
Changes
Added
p_significance()
methods for frequentist models.Methods for
degrees_of_freedom()
have been removed.degrees_of_freedom()
now callsinsight::get_df()
.model_parameters()
for data frames anddraws
objects from package posterior also gets anexponentiate
argument.
parameters 0.22.1
CRAN release: 20240721
Breaking changes
 Revised calculation of the second generation pvalue (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.
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()
foranova()
from mixed models now also includes the denominator degrees of freedom in the output (df_error
).print(..., pretty_names = "labels")
for tobitmodels 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 (oldrelease).
parameters 0.22.0
CRAN release: 20240620
Breaking changes
Deprecated arguments in
model_parameters()
forhtest
,aov
andBFBayesFactor
objects were removed.Argument
effectsize_type
is deprecated. Please usees_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()
forfactor_analysis()
.Fixed wrong ORCID in
DESCRIPTION
.
parameters 0.21.7
CRAN release: 20240514
Bug fixes
Fixes issue in
compare_parameters()
for models from package blme.Fixed conflict in
model_parameters()
when bothinclude_reference = TRUE
andpretty_names = "labels"
were used. Now, pretty labels are correctly updated and preserved.
parameters 0.21.6
CRAN release: 20240318
Changes
include_reference
can now directly be set toTRUE
inmodel_parameters()
and doesn’t require a call toprint()
anymore.compare_parameters()
gains ainclude_reference
argument, to add the reference category of categorical predictors to the parameters table.print_md()
forcompare_parameters()
now by default uses the tinytable package to create markdown tables. This allows better control for column heading spanning over multiple columns.
Bug fixes
Fixed issue with parameter names for
model_parameters()
and objects from package epiR.Fixed issue with
exponentiate = TRUE
formodel_parameters()
with models of classclmm
(package ordinal), when model had nocomponent
column (e.g., no scale or location parameters were returned).include_reference
now also works when factor were created “onthefly” inside the model formula (i.e.y ~ as.factor(x)
).
parameters 0.21.4
CRAN release: 20240203
Breaking changes
 The
exponentiate
argument ofmodel_parameters()
formarginaleffects::predictions()
now defaults toFALSE
, in line with all the othermodel_parameters()
methods.
Changes
model_parameters()
for models of package survey now gives informative messages whenbootstrap = TRUE
(which is currently not supported).n_factors()
now also returns the explained variance for the number of factors as attributes.model_parameters()
for objects of package metafor now warns when unsupported arguments (likevcov
) are used.Improved documentation for
pool_parameters()
.
Bug fixes
print(include_reference = TRUE)
formodel_parameters()
did not work when run inside a pipechain.Fixed issues with
format()
for objects returned bycompare_parameters()
that included mixed models.
parameters 0.21.3
CRAN release: 20231102
Changes
principal_components()
andfactor_analysis()
now also work when argumentn = 1
.print_md()
forcompare_parameters()
now gains more arguments, similar to theprint()
method.bootstrap_parameters()
andmodel_parameters()
now accept bootstrapped samples returned bybootstrap_model()
.The
print()
method formodel_parameters()
now also yields a warning for models with logitlinks when possible issues with (quasi) complete separation occur.
Bug fixes
Fixed issue in
print_html()
for objects from package ggeffects.Fixed issues for
nnet::multinom()
with wideformat response variables (usingcbind()
).Minor fixes for
print_html()
method formodel_parameters()
.Robust standard errors (argument
vcov
) now works forplm
models.
parameters 0.21.2
CRAN release: 20230916
Changes
Minor improvements to factor analysis functions.
The
ci_digits
argument of theprint()
method formodel_parameters()
now defaults to the same value ofdigits
.model_parameters()
for objects from package marginaleffects now also accepts theexponentiate
argument.The
print()
,print_html()
,print_md()
andformat()
methods formodel_parameters()
get aninclude_reference
argument, to add the reference category of categorical predictors to the parameters table.
Bug fixes
Fixed issue with wrong calculation of teststatistic and pvalues in
model_parameters()
forfixest
models.Fixed issue with wrong column header for
glm
models withfamily = binomial("identiy")
.Minor fixes for
dominance_analysis()
.
parameters 0.21.1
CRAN release: 20230526
Changes to functions
model_parameters()
now also prints correct “pretty names” when predictors where converted to ordered factors inside formulas, e.g.y ~ as.ordered(x)
.model_parameters()
now prints a message when thevcov
argument is provided andci_method
is explicitly set to"profile"
. Else, whenvcov
is notNULL
andci_method
isNULL
, it defaults to"wald"
, to return confidence intervals based on robust standard errors.
parameters 0.21.0
CRAN release: 20230419
parameters 0.20.3
CRAN release: 20230405
General
Added support for models of class
hglm
(hglm),mblogit
(mclogit),fixest_multi
(fixest), andphylolm
/phyloglm
(phylolm).as.data.frame
methods for extracting posterior draws viabootstrap_model()
have been retired. Instead, directly usingbootstrap_model()
is recommended.
Changes to functions
equivalence_test()
gets a method forggeffects
objects from package ggeffects.equivalence_test()
now prints theSGPV
column instead of% in ROPE
. This is because the former% in ROPE
actually was equivalent to the second generation pvalue (SGPV) and refers to the proportion of the range of the confidence interval that is covered by the ROPE. However,% in ROPE
did not refer to the probability mass of the underlying distribution of a confidence interval that was covered by the ROPE, hence the old column name was a bit misleading.Fixed issue in
model_parameters.ggeffects()
to address forthcoming changes in the ggeffects package.
Bug fixes
When an invalid or not supported value for the
p_adjust
argument inmodel_parameters()
is provided, the valid options were not shown in correct capital letters, where appropriate.Fixed bug in
cluster_analysis()
forinclude_factors = TRUE
.Fixed warning in
model_parameters()
andci()
for models from package glmmTMB whenci_method
was either"profile"
or"uniroot"
.
parameters 0.20.2
CRAN release: 20230127
General
Reduce unnecessary warnings.
The deprecated argument
df_method
inmodel_parameters()
was removed.Output from
model_parameters()
for objects returned bymanova()
andcar::Manova()
is now more consistent.
Bug fix
Fixed issues in tests for
mmrm
models.Fixed issue in
bootstrap_model()
for models of classglmmTMB
with dispersion parameters.Fixed failing examples.
parameters 0.20.1
CRAN release: 20230111
Changes
model_parameters()
now includes aGroup
column forstanreg
orbrmsfit
models with random effects.The
print()
method formodel_parameters()
now uses the same pattern to print random effect variances for Bayesian models as for frequentist models.
Bug fix
Fixed issue with the
print()
method forcompare_parameters()
, which duplicated random effects parameters rows in some edge cases.Fixed issue with the
print()
method forcompare_parameters()
, which didn’t work properly whenci=NULL
.
parameters 0.20.0
CRAN release: 20221121
Breaking
The deprecated argument
df_method
inmodel_parameters()
is now defunct and throws an error when used.The deprecated functions
ci_robust()
,p_robust()
andstandard_error_robust
have been removed. These were superseded by thevcov
argument inci()
,p_value()
, andstandard_error()
, respectively.The
style
argument incompare_parameters()
was renamed intoselect
.
New functions
p_function()
, to print and plot pvalues and compatibility (confidence) intervals for statistical models, at different levels. This allows to see which estimates are most compatible with the model at various compatibility levels.p_calibrate()
, to compute calibrated pvalues.
Changes
model_parameters()
andcompare_parameters()
now use the unicode character for the multiplicationsign as interaction mark (i.e.\u00d7
). Useoptions(parameters_interaction = <value>)
or the argumentinteraction_mark
to use a different character as interaction mark.The
select
argument incompare_parameters()
, which is used to control the table column elements, now supports an experimental gluelike syntax. See this vignette Printing Model Parameters. Furthermore, theselect
argument can also be used in theprint()
method formodel_parameters()
.print_html()
gets afont_size
andline_padding
argument to tweak the appearance of HTML tables. Furthermore, argumentsselect
andcolumn_labels
are new, to customize the column layout of tables. See examples in?display
.Consolidation of vignettes on standardization of model parameters.
Minor speed improvements.
Bug fix
model_parameters().BFBayesFactor
no longer drops theBF
column if the Bayes factor isNA
.The
print()
anddisplay()
methods formodel_parameters()
from Bayesian models now pass the...
toinsight::format_table()
, allowing extra arguments to be recognized.Fixed footer message regarding the approximation method for CU and pvalues for mixed models.
Fixed issues in the
print()
method forcompare_parameters()
with mixed models, when some models contained withinbetween components (seewb_component
) and others did not.
parameters 0.19.0
CRAN release: 20221005
Breaking
Arguments that calculate effectsize in
model_parameters()
forhtest
, Anova objects and objects of classBFBayesFactor
were revised. Instead of single arguments for the different effectsizes, there is now one argument,effectsize_type
. The reason behind this change is that meanwhile many new type of effectsizes have been added to the effectsize package, and the generic argument allows to make use of those effect sizes.The attribute name in PCA / EFA has been changed from
data_set
todataset
.The minimum needed R version has been bumped to
3.6
.Removed deprecated argument
parameters
frommodel_parameters()
.standard_error_robust()
,ci_robust()
andp_value_robust()
are now deprecated and superseded by thevcov
andvcov_args
arguments in the related methodsstandard_error()
,ci()
andp_value()
, respectively.Following functions were moved from package parameters to performance:
check_sphericity_bartlett()
,check_kmo()
,check_factorstructure()
andcheck_clusterstructure()
.
Changes to functions
Added
sparse
option toprincipal_components()
for sparse PCA.The
pretty_names
argument from theprint()
method can now also be"labels"
, which will then use variable and value labels (if data is labelled) as pretty names. If no labels were found, default pretty names are used.bootstrap_model()
for models of classglmmTMB
andmerMod
gains acluster
argument to specify optional clusters when theparallel
option is set to"snow"
.Pvalue adjustment (argument
p_adjust
inmodel_parameters()
) is now performed after potential parameters were removed (usingkeep
ordrop
), so adjusted pvalues is only applied to the parameters of interest.Robust standard errors are now supported for
fixest
models with thevcov
argument.print()
formodel_parameters()
gains afooter
argument, which can be used to suppress the footer in the output. Further more, iffooter = ""
orfooter = FALSE
inprint_md()
, no footer is printed.simulate_model()
andsimulate_parameters()
now pass...
toinsight::get_varcov()
, to allow simulated draws to be based on heteroscedasticity consistent variance covariance matrices.The
print()
method forcompare_parameters()
was improved for models with multiple components (e.g., mixed models with fixed and random effects, or models with count and zeroinflation parts). For these models,compare_parameters(effects = "all", component = "all")
prints more nicely.
parameters 0.18.2
CRAN release: 20220810
New functions
 New function
dominance_analysis()
, to compute dominance analysis statistics and designations.
Changes to functions
 Argument
ci_random
inmodel_parameters()
defaults toNULL
. It uses a heuristic to determine if random effects confidence intervals are likely to take a long time to compute, and automatically includes or excludes those confidence intervals. Setci_random
toTRUE
orFALSE
to explicitly calculate or omit confidence intervals for random effects.
Bug fixes
Fix issues in
pool_parameters()
for certain models with special components (likeMASS::polr()
), that failed when argumentcomponent
was set to"conditional"
(the default).Fix issues in
model_parameters()
for multiple imputation models from package Hmisc.
parameters 0.18.1
CRAN release: 20220529
General
It is now possible to hide messages about CI method below tables by specifying
options("parameters_cimethod" = FALSE)
(#722). By default, these messages are displayed.model_parameters()
now supports objects from package marginaleffects and objects returned bycar::linearHypothesis()
.Added
predict()
method tocluster_meta
objects.Reorganization of docs for
model_parameters()
.
Changes to functions
model_parameters()
now also includes standard errors and confidence intervals for slopeslopecorrelations of random effects variances.model_parameters()
for mixed models gains aci_random
argument, to toggle whether confidence intervals for random effects parameters should also be computed. Set toFALSE
if calculation of confidence intervals for random effects parameters takes too long.ci()
for glmmTMB models withmethod = "profile"
is now more robust.
Bug fixes
Fixed issue with glmmTMB models when calculating confidence intervals for random effects failed due to singular fits.
display()
now correctly includes custom text and additional information in the footer (#722).Fixed issue with argument
column_names
incompare_parameters()
when strings contained characters that needed to be escaped for regular expressions.Fixed issues with unknown arguments in
model_parameters()
for lavaan models whenstandardize = TRUE
.
parameters 0.18.0
CRAN release: 20220524
Breaking Changes

model_parameters()
now no longer treats data frame inputs as posterior samples. Rather, for data frames, nowNULL
is returned. If you want to treat a data frame as posterior samples, set the new argumentas_draws = TRUE
.
New functions
sort_parameters()
to sort model parameters by coefficient values.standardize_parameters()
,standardize_info()
andstandardise_posteriors()
to standardize model parameters.
Changes to functions
model_parameters()
model_parameters()
for mixed models from package lme4 now also reports confidence intervals for random effect variances by default. Formerly, CIs were only included whenci_method
was"profile"
or"boot"
. The merDeriv package is required for this feature.model_parameters()
forhtest
objects now also supports models fromvar.test()
.Improved support for
anova.rms
models inmodel_parameters()
.model_parameters()
now supportsdraws
objects from package posterior anddeltaMethods
objects from package car.model_parameters()
now checks arguments and informs the user if specific given arguments are not supported for that model class (e.g.,"vcov"
is currently not supported for models of class glmmTMB).
Bug fixes
The
vcov
argument, used for computing robust standard errors, did not calculate the correct pvalues and confidence intervals for models of classlme
.pool_parameters()
did not save all relevant model information as attributes.model_parameters()
for models from package glmmTMB did not work whenexponentiate = TRUE
and model contained a dispersion parameter that was different than sigma. Furthermore, exponentiating falsely exponentiated the dispersion parameter.
parameters 0.17.0
CRAN release: 20220310
General

Added options to set defaults for different arguments. Currently supported:

options("parameters_summary" = TRUE/FALSE)
, which sets the default value for thesummary
argument inmodel_parameters()
for nonmixed models. 
options("parameters_mixed_summary" = TRUE/FALSE)
, which sets the default value for thesummary
argument inmodel_parameters()
for mixed models.

Minor improvements for
print()
methods.
Robust uncertainty estimates:
 The
vcov_estimation
,vcov_type
, androbust
arguments are deprecated in these functions:model_parameters()
,parameters()
,standard_error()
,p_value()
, andci()
. They are replaced by thevcov
andvcov_args
arguments.  The
standard_error_robust()
andp_value_robust()
functions are superseded by thevcov
andvcov_args
arguments of thestandard_error()
andp_value()
functions.  Vignette: https://easystats.github.io/parameters/articles/model_parameters_robust.html
 The
Bug fixes
Fixed minor issues and edge cases in
n_clusters()
and related cluster functions.Fixed issue in
p_value()
that returned wrong pvalues forfixest::feols()
.
parameters 0.16.0
CRAN release: 20220112
General
Improved speed performance for
model_parameters()
, in particular for glm’s and mixed models where random effect variances were calculated.Added more options for printing
model_parameters()
. See also revised vignette: https://easystats.github.io/parameters/articles/model_parameters_print.html
Changes to functions
model_parameters()
model_parameters()
for mixed models gains aninclude_sigma
argument. IfTRUE
, adds the residual variance, computed from the random effects variances, as an attribute to the returned data frame. Including sigma was the default behaviour, but now defaults toFALSE
and is only included wheninclude_sigma = TRUE
, because the calculation was very time consuming.model_parameters()
formerMod
models now also computes CIs for the random SD parameters whenci_method="boot"
(previously, this was only possible whenci_method
was"profile"
).model_parameters()
forglmmTMB
models now computes CIs for the random SD parameters. Note that these are based on a Waldzdistribution.Similar to
model_parameters.htest()
, themodel_parameters.BFBayesFactor()
method gainscohens_d
andcramers_v
arguments to control if you need to add frequentist effect size estimates to the returned summary data frame. Previously, this was done by default.Column name for coefficients from emmeans objects are now more specific.
model_prameters()
forMixMod
objects (package GLMMadaptive) gains arobust
argument, to compute robust standard errors.
Bug fixes
Fixed bug with
ci()
for classmerMod
whenmethod="boot"
.Fixed issue with correct association of components for ordinal models of classes
clm
andclm2
.Fixed issues in
random_parameters()
andmodel_parameters()
for mixed models without random intercept.Confidence intervals for random parameters in
model_parameters()
failed for (some?)glmer
models.Fix issue with default
ci_type
incompare_parameters()
for Bayesian models.
parameters 0.15.0
CRAN release: 20211018
Breaking changes

Following functions were moved to the new datawizard package and are now reexported from parameters package:
center()
convert_data_to_numeric()
data_partition()
demean()
(and its aliasesdegroup()
anddetrend()
)smoothness()
Note that these functions will be removed in the next release of parameters package and they are currently being reexported only as a convenience for the package developers. This release should provide them with time to make the necessary changes before this breaking change is implemented.

Following functions were moved to the performance package:
check_heterogeneity()
General
The handling to approximate the degrees of freedom in
model_parameters()
,ci()
andp_value()
was revised and should now be more consistent. Some bugs related to the previous computation of confidence intervals and pvalues have been fixed. Now it is possible to change the method to approximate degrees of freedom for CIs and pvalues using theci_method
, resp.method
argument. This change has been documented in detail in?model_parameters
, and online here: https://easystats.github.io/parameters/reference/model_parameters.htmlMinor changes to
print()
for glmmTMB with dispersion parameter.Added vignette on printing options for model parameters.
Changes to functions
model_parameters()
The
df_method
argument inmodel_parameters()
is deprecated. Please useci_method
now.model_parameters()
withstandardize = "refit"
now returns random effects from the standardized model.model_parameters()
andci()
forlmerMod
models gain a"residuals"
option for theci_method
(resp.method
) argument, to explicitly calculate confidence intervals based on the residual degrees of freedom, when present.model_parameters()
supports following new objects:trimcibt
,wmcpAKP
,dep.effect
(in WRS2 package),systemfit
model_parameters()
gains a new argumenttable_wide
for ANOVA tables. This can be helpful for users who may wish to report ANOVA table in wide format (i.e., with numerator and denominator degrees of freedom on the same row).model_parameters()
gains two new arguments,keep
anddrop
.keep
is the new names for the formerparameters
argument and can be used to filter parameters. Whilekeep
selects those parameters whose names match the regular expression pattern defined inkeep
,drop
is the counterpart and excludes matching parameter names.When
model_parameters()
is called withverbose = TRUE
, andci_method
is not the default value, the printed output includes a message indicating which approximationmethod for degrees of freedom was used.model_parameters()
for mixed models withci_method = "profile
computes (profiled) confidence intervals for both fixed and random effects. Thus,ci_method = "profile
allows to add confidence intervals to the random effect variances.model_parameters()
should longer fail for supported model classes when robust standard errors are not available.
Other functions
n_factors()
the methods based on fit indices have been fixed and can be included separately (package = "fit"
). Also added an_max
argument to crop the output.compare_parameters()
now also accepts a list of model objects.describe_distribution()
getsverbose
argument to toggle warnings and messages.format_parameters()
removes dots and underscores from parameter names, to make these more “human readable”.The experimental calculation of pvalues in
equivalence_test()
was replaced by a proper calculation pvalues. The argumentp_value
was removed and pvalues are now always included.Minor improvements to
print()
,print_html()
andprint_md()
.
Bug fixes
The random effects returned by
model_parameters()
mistakenly displayed the residuals standard deviation as squareroot of the residual SD.Fixed issue with
model_parameters()
for brmsfit objects that model standard errors (i.e. for metaanalysis).Fixed issue in
model_parameters
forlmerMod
models that, by default, returned residual degrees of freedom in the statistic column, but confidence intervals were based onInf
degrees of freedom instead.Fixed issue in
ci_satterthwaite()
, which usedInf
degrees of freedom instead of the Satterthwaite approximation.Fixed issue in
model_parameters.mlm()
when model contained interaction terms.Fixed issue in
model_parameters.rma()
when model contained interaction terms.Fixed sign error for
model_parameters.htest()
for objects created witht.test.formula()
(issue #552)Fixed issue when computing random effect variances in
model_parameters()
for mixed models with categorical random slopes.
parameters 0.14.0
CRAN release: 20210529
Breaking changes
check_sphericity()
has been renamed intocheck_sphericity_bartlett()
.Removed deprecated arguments.
model_parameters()
for bootstrapped samples used in emmeans now treats the bootstrap samples as samples from posterior distributions (Bayesian models).
New supported model classes

SemiParBIV
(GJRM),selection
(sampleSelection),htest
from the survey package,pgmm
(plm).
New functions
 Added a
summary()
method formodel_parameters()
, which is a convenient shortcut forprint(..., select = "minimal")
.
Changes to functions
model_parameters()
model_parameters()
gains aparameters
argument, which takes a regular expression as string, to select specific parameters from the returned data frame.print()
formodel_parameters()
andcompare_parameters()
gains agroups
argument, to group parameters in the output. Furthermore,groups
can be used directly as argument inmodel_parameters()
andcompare_parameters()
and will be passed to theprint()
method.model_parameters()
for ANOVAs now saves the type as attribute and prints this information as footer in the output as well.model_parameters()
for htestobjects now saves the alternative hypothesis as attribute and prints this information as footer in the output as well.model_parameters()
passes argumentstype
,parallel
andn_cpus
down tobootstrap_model()
whenbootstrap = TRUE
.
other
bootstrap_models()
for merMod and glmmTMB objects gains further arguments to set the type of bootstrapping and to allow parallel computing.bootstrap_parameters()
gains theci_method
type"bci"
, to compute biascorrected and accelerated bootstrapped intervals.ci()
forsvyglm
gains amethod
argument.
Bug fixes
Fixed issue in
model_parameters()
for emmGrid objects with Bayesian models.Arguments
digits
,ci_digits
andp_digits
were ignored forprint()
and only worked when used in the call tomodel_parameters()
directly.
parameters 0.13.0
CRAN release: 20210408
General
 Revised and improved the
print()
method formodel_parameters()
.
New supported model classes

blrm
(rmsb),AKP
,med1way
,robtab
(WRS2),epi.2by2
(epiR),mjoint
(joineRML),mhurdle
(mhurdle),sarlm
(spatialreg),model_fit
(tidymodels),BGGM
(BGGM),mvord
(mvord)
Changes to functions
model_parameters()
model_parameters()
forblavaan
models is now fully treated as Bayesian model and thus relies on the functions from bayestestR (i.e. ROPE, Rhat or ESS are reported) .The
effects
argument frommodel_parameters()
for mixed models was revised and now shows the random effects variances by default (same functionality asrandom_parameters()
, but mimicking the behaviour frombroom.mixed::tidy()
). When thegroup_level
argument is set toTRUE
, the conditional modes (BLUPs) of the random effects are shown.model_parameters()
for mixed models now returns anEffects
column even when there is just one type of “effects”, to mimic the behaviour frombroom.mixed::tidy()
. In conjunction withstandardize_names()
users can get the same column names as intidy()
formodel_parameters()
objects.model_parameters()
for ttests now uses the group values as column names.print()
formodel_parameters()
gains azap_small
argument, to avoid scientific notation for very small numbers. Instead,zap_small
forces to round to the specified number of digits.To be internally consistent, the degrees of freedom column for
lqm(m)
andcgam(m)
objects (with tstatistic) is calleddf_error
.model_parameters()
gains asummary
argument to add summary information about the model to printed outputs.Minor improvements for models from quantreg.
model_parameters
supports rankbiserial, rank epsilonsquared, and Kendall’s W as effect size measures forwilcox.test()
,kruskal.test
, andfriedman.test
, respectively.
Other functions

describe_distribution()
gets aquartiles
argument to include 25th and 75th quartiles of a variable.
Bug fixes
Fixed issue with noninitialized argument
style
indisplay()
forcompare_parameters()
.Make
print()
forcompare_parameters()
work with objects that have “simple” column names for confidence intervals with missing CIlevel (i.e. when column is named"CI"
instead of, say,"95% CI"
).Fixed issue with
p_adjust
inmodel_parameters()
, which did not work for adjustmentmethods"BY"
and"BH"
.Fixed issue with
show_sigma
inprint()
formodel_parameters()
.Fixed issue in
model_parameters()
with incorrect order of degrees of freedom.
parameters 0.12.0
CRAN release: 20210221
General
Rollback R dependency to R >= 3.4.

Bootstrapped estimates (from
bootstrap_model()
orbootstrap_parameters()
) can be passed toemmeans
to obtain bootstrapped estimates, contrasts, simple slopes (etc) and their CIs. These can then be passed to
model_parameters()
and related functions to obtain standard errors, pvalues, etc.
 These can then be passed to
Breaking changes
model_parameters()
now always returns the confidence level for as additionalCI
column.The
rule
argument inequivalenct_test()
defaults to"classic"
.
New supported model classes

crr
(cmprsk),leveneTest()
(car),varest
(vars),ergm
(ergm),btergm
(btergm),Rchoice
(Rchoice),garch
(tseries)
New functions

compare_parameters()
(and its aliascompare_models()
) to show / print parameters of multiple models in one table.
Changes to functions
Estimation of bootstrapped pvalues has been rewritten to be more accurate.
model_parameters()
for mixed models gains aneffects
argument, to return fixed, random or both fixed and random effects parameters.Revised printing for
model_parameters()
for metafor models.model_parameters()
for metafor models now recognized confidence levels specified in the function call (via argumentlevel
).Improved support for effect sizes in
model_parameters()
from anova objects.
Bug fixes
Fixed edge case when formatting parameters from polynomial terms with many degrees.
Fixed issue with random sampling and dropped factor levels in
bootstrap_model()
.