File: NEWS.md

package info (click to toggle)
r-cran-parameters 0.11.0-1
  • links: PTS, VCS
  • area: main
  • in suites: bullseye, sid
  • size: 3,044 kB
  • sloc: sh: 15; makefile: 2
file content (541 lines) | stat: -rw-r--r-- 18,091 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
# parameters 0.10.2

## New supported model classes

* `coxr` (*coxrobust*), `coeftest` (*lmtest*), `ivfixed` (*ivfixed*), 
  `ivprobit` (*ivprobit*), `riskRegression` (*riskRegression*),
  `fitdistr` (*MASS*), `yuen`, `t1way`, `onesampb`, `mcp1` and `mcp2` (*WRS2*),
  `Anova.mlm` (*car*), `rqs` (*quantreg*), `lmodel2` (*lmodel2*),
  `summary.lm`, `PMCMR`, `osrt` and `trendPMCMR` (*PMCMRplus*), `bamlss`
  (*bamlss*).

## New functions

### Printing and table Formatting

* `print_html()` as an alias for `display(format = "html")`. This allows to
  print tabular outputs from data frames (as returned by most functions in
  _parameters_) into nicely rendered HTML markdown tables.

## Changes to functions

* Added more effect size measures to `model_parameters()` for `htest` objects.

* `model_parameters()` for anova objects gains a `power` argument, to calculate
  the power for each parameter.

* `ci()` for models from *lme4* and *glmmTMB* can now computed profiled confidence
  intervals, using `method = "profile"`. Consequently, `model_parameters()` with
  `df_method = "profile"` also computes profiled confidence intervals. For
  models of class `glmmTMB`, option `"uniroot"` is also available.

## Bug fixes

* `model_parameters()` for t-tests when `standardize_d = TRUE`, did not return
  columns for the group-specific means.

* Fixed issue in `p_value()` for `fixest::feols()`.

* Fixed issue in `model_parameters()` for `glmer()` models with p-values that
  were calculated with `df_method = "ml1"` or `df_method = "betwithin"`.

* Fixed issue in `model_parameters()` for multinomial models when response
  was a character vector (and no factor).

* Fixed issue in `print_md()` for model-parameters objects from Bayesian models.

* Fixed issues with printing of model parameters for multivariate response 
  models from *brms*.
  
* Fixed issue with paired t-tests and `model_parameters()`.

# parameters 0.10.1

## New functions

* `format_p_adjust()`, to create pretty names for p-adjustment methods.

## Bug fixes

* Fixed breaking code / failing tests due to latest _effectsize_ update.

* Fixed issue with `model_parameters()` for models of class `mlm`.

* Undocumented arguments `digits`, `ci_digits` and `p_digits` worked for 
  `print()`, but not when directly called inside `model_parameters()`. Now,
  `model_parameters(model, digits = 5, ci_digits = 8)` works again.

* Fixed some minor printing-issues.

# parameters 0.10.0

## Breaking changes

* The default-method for effect sizes in `model_parameters()` for Anova-models
  (i.e. when arguments `omega_squared`, `eta_squared` or `epsilon_squared` are
  set to `TRUE`) is now `"partial"`, as initially intended.

* Column names for degrees of freedom were revised. `"df_residual"` was replaced
  by the more generic `"df_error"`. Moreover, models of class `htest` now also
  have the column name `"df_error"` and no longer `"df"` (where applicable).

* Some re-exports for functions that were moved to *insight* longer ago, were
  now removed.

## New supported model classes

* `Glm` (*rms*), `mediate` (*mediation*).

* `model_parameters()` supports `Gam` models (*gam*), `ridgelm` (*MASS*), 
  `htest` objects from `oneway.test()`, `chisq.test()`, `prop.test()`,
  `mcnemar.test()` and `pairwise.htest` objects, `mcmc.list` (e.g. from
  *bayesGARCH*).

## New functions

### Printing and table Formatting

* `display()`, to format output from package-functions into different formats.

* `print_md()` as an alias for `display(format = "markdown")`. This allows to
  print tabular outputs from data frames (as returned by most functions in
  _parameters_) into nicely rendered markdown tables.

* `format()`, to create a "pretty data frame" with nicer column names and
  formatted values. This is one of the worker-functions behind `print()` or 
  `print_md()`.

## Changes to functions

### `model_parameters()`

* `model_parameters()` for Anova-models (of class `aov`, `anova` etc.) gains a
  `ci`-argument, to add confidence intervals to effect size parameters.

* `model_parameters()` for `htest` objects gains a `cramers_v` and `phi`
  argument, to compute effect size parameters for objects from `chisq.test()`,
  and a `standardized_D` argument, to compute effect size parameters for 
  objects from `t.test()`.

* `model_parameters()` for `metafor`-models is more stable when called from inside
  functions.
  
* `model_parameters()` for *metaBMA*-models now includes prior information for
  the meta-parameters.

* `model_parameters()` for meta-analysis-models gains a `include_studies`-argument,
  to include or remove studies from the output.
  
* `model_parameters()` for gam-models now includes the residual df for smooth
   terms, and no longer the reference df.
  
* Slightly revised and improved the `print()` method for `model_parameters()`.

### Other functions

* `describe_distribution()` now includes the name of the centrality index in the
  `CI`-column, when `centrality = "all"`.

* `pool_parameters()` gains a `details`-argument. For mixed models, and if
  `details = TRUE`, random effect variances will also be pooled.

## Bug fixes

* Fixed issue in `ci()` for *lme* models with non-positive definite
  variance-covariance.
  
* Fixed issue in `model_parameters()` for `nnet::multinom()`, `lqmm::lqm()`,
  `mgcv::gam()`, and `margins::margins()` models, and models from package 
  *blme*.


# parameters 0.9.0

## New supported model classes

* Support for `maov` (*stats*), `HLfit` (*spaMM*), `scam` (*scam*), preliminary
  support for `emm_list` (*emmeans*), `merModList` (*merTools*), `meta_random`,
  `meta_bma` and `meta_fixed` (*metaBMA*).

## New functions

* `pool_parameters()`, to pool parameters estimates from multiple models.

* `degroup()`, as a more generic case for `demean()`.

* `center()`, to center variables.

## General

* Better support for (weighted) multivariate response models of class `mlm` for
  functions like `model_parameters()` or `simulate_parameters()`.

* `standardize_names()` is now re-exported from the *insight* package.

## Changes to functions

### Printing model parameters

* `print()` for `model_parameters()` now names the coefficients column depending
  on the model type (i.e. `"Odds Ratios"` for logistic regression when
  `exponentiate = TRUE` etc.)

* `print()` for `model_parameters()` gains a `show_sigma` argument, to show or
  hide information on the residual standard deviation.

* `print()` for `model_parameters()` displays a message for Bayesian models,
  indicating which method to compute credible intervals was used.

### Other changes

* `data_partition()` gets a `seed` argument, to explicitly set the seed before
  random sampling of test and training data.

* Revised `parameters_table()`, to improve readability of printed output.

## Bug fixes

* Fixed issues in `model_parameters()` for *vgam* and *mira* objects.

* Fixed issue where `model_parameters()` for *emmGrid* objects falsely removed
  the `Coefficient` column.

* Fixed issue in `parameters_type()` for factors with different effects-coding
  than treatment contrasts.

* Fixed issues due to latest _effectsize_ update.

# parameters 0.8.6

## Bug fixes

* Fixed issues with *glmmTMB* models with dispersion-parameter.

* Fixed issue where `model_parameters()` for *glmmTMB* models falsely removed
  the `Component` column.

* Fixed issue with missing CI columns in `model_parameters()` when `standardize`
  was one of the options except `"refit"`.

* `parameters_type()` did not correctly detect interaction terms for specific
  patterns like `scale()` included in the interaction.

# parameters 0.8.5

## General

* Added vignette on model parameters and missing data.

* Update citation.

## New supported model classes

* Support for `mipo` (*mice*), `lqm` and `lqmm` (*lqmm*). Preliminary support
  for `semLME` (*smicd*), `mle2` (*bbmle*), `mle` (*stats4*)

* `model_parameters()` for objects of class `mira` (*mice*).

## Changes to functions

* `model_parameters()` gets a specific behaviour for brms-meta-analysis models.

* `model_parameters()` for *lavaan* and *blavaan* now also prints self-defined
  parameters.

* `model_parameters()` for *lavaan* and *blavaan* gains more option for
  standardized parameters.

## Bug fixes

* Fix issue in `model_parameters()` for `coxph.penal` models.

* Fix issue in `model_parameters.metaplus()` with random effects.

* Fix issue in `check_heterogeneity()` when `x` was a mixed model.

* Fix issue in `check_heterogeneity()` for data with missing values.

* Fix issue in `dof_ml1()` when random-effect terms where character vectors.

* Fix issue in `print()` method for `model_parameters()` that printed empty
  lines for rows with complete missing values. Empty lines are now removed.

* Fix issue in `parameters_type()` when `exp()` was used in a model formula.

# parameters 0.8.2

## New supported models

* `metaplus` (*metaplus*), `glht` (*multcomp*), `glmm` (*glmm*), `manova`
  (*stats*), `crq` and `crqs` (*quantreg*)

* Improved support for models from the *rms* package.

## Changes to functions

* Improved parameters formatting for ordered factors in `model_parameters()`
  (and `format_parameters()`).

* Argument `df_method` can now also be applied to GLMs, to allow calculation of
  confidence intervals based on Wald-approximation, not profiled confidence
  intervals. This speeds up computation of CIs for models fit to large data
  sets.

* Improved `select_parameters()` for mixed models, and revised docs and
  associated vignette.

## Bug fixes

* Allow `threshold` to be passed to `efa_to_cfa()` when the model is from
  `factor_analysis()`.

* Allow correlation matrix to be passed to `factor_analysis()`.

* Fix CRAN check issues.

* Fix issue in `model_parameters()` for models with non-estimable parameters or
  statistics.

* Fix issue in `model_parameters()` for *plm* models with only one parameter.

* Fix issue in `check_heterogeneity()` in case no predictor would cause
  heterogeneity bias.

* Make sure *clubSandwich* is used conditionally in all places, to properly pass
  CRAN checks.

# parameters 0.8.0

## New supported models

* `robmixglm` (*robmixglm*), `betaor`, `betamfx`, `logitor`, `poissonirr`,
  `negbinirr`, `logitmfx`, `probitmfx`, `poissonmfx`, `negbinmfx` (*mfx*),
  partial support `emmGrid` (*emmeans*)

## Changes to functions

### `simulate_parameters()` and `simulate_model()`

* has a nicer `print()` method.

* now also simulate parameters from the dispersion model for *glmmTMB* objects.

* gets a `verbose` argument, to show or hide warnings and messages.

## Bug fixes

* fix issue with rank deficient models.

# parameters 0.7.0

## General

* We changed the computation of confidence intervals or standard errors, so
  these are now based on a t-distribution with degrees of freedom and not normal
  distribution assuming infinite degrees of freedom. This was implemented for
  most functions before and only affects few functions (like
  `equivalence_test()` or CIs for standardized parameters from
  `model_parameters()` when standardization method was `"posthoc"`).

## New supported models

* `averaging` (*MuMIn*), `bayesx` (*R2BayesX*), `afex_aov` (*afex*)

## New functions

* `check_heterogeneity()` as a small helper to find variables that have a
  within- and between-effect related to a grouping variable (and thus, may
  result in heterogeneity bias, see [this
  vignette](https://easystats.github.io/parameters/articles/demean.html)).

## Changes to functions

### `equivalence_test()`

* gains a `rule` argument, so equivalence testing can be based on different
  approaches.

* for mixed models gains an `effect` argument, to perform equivalence testing on
  random effects.

* gains a `p_values` argument, to calculate p-values for the equivalence test.

* now supports more frequentist model objects.

### `describe_distribution()`

* now works on grouped data frames.

* gains `ci` and `iterations` arguments, to compute confidence intervals based
  on bootstrapping.

* gains a `iqr` argument, to compute the interquartile range.

* `SE` column was removed.

### `model_parameters()`

* `model_parameters()` for Stan-models (*brms*, *rstanarm*) gains a
  `group_level` argument to show or hide parameters for group levels of random
  effects.

* Improved accuracy of confidence intervals in `model_parameters()` with
  `standardize = "basic"` or `standardize = "posthoc"`.

* `model_parameters.merMod()` no longer passes `...` down to bootstrap-functions
  (i.e. when `bootstrap = TRUE`), as this might conflict with
  `lme4::bootMer()`.

* For ordinal models (like `MASS::polr()` or `ordinal::clm()`), a `Component`
  column is added, indicating intercept categories (`"alpha"`) and estimates
  (`"beta"`).

* The `select`-argument from `print.parameters_model()` now gets a
  `"minimal"`-option as shortcut to print coefficients, confidence intervals and
  p-values only.

### Other changes

* `parameters_table()` and `print.parameters_model()` now explicitly get
  arguments to define the digits for decimal places used in output.

* `ci()`, `standard_error()`, `p_value()` and `model_parameters()` for *glmmTMB*
  models now also works for dispersion models.

## Bug fixes

* Fixed issue in `equivalence_test()` for mixed models.

* Fixed bug for `model_parameters.anova(..., eta_squared = "partial")` when
  called with non-mixed models.

* Fixed issue with wrong degrees of freedom in `model_parameters()` for *gam*
  models.

* Fixed issue with unused arguments in `model_parameters()`.

# parameters 0.6.1

## General

* Remove 'Zelig' from suggested packages, as it was removed from CRAN.

## Changes to functions

### model_parameters()

* `model_parameters()` now also transforms standard errors when `exponentiate =
  TRUE`.

* `model_parameters()` for `anova()` from mixed models can now also compute
  effect sizes like eta squared.

* `model_parameters()` for `aov()` gains a `type`-argument to compute type-1,
  type-2 or type-3 sums of squares.

* `model_parameters()` for Bayesian models gains a `standardize` argument, to
  return standardized parameters from the posterior distribution.

* Improved `print()` method for `model_parameters()` for nested `aov()`
  (repeated measurements).

* You can now control whether `demean()` should add attributes to indicate
  within- and between-effects. This is only relevant for the `print()`-method of
  `model_parameters()`.

## Bug fixes

* Fixed `model_parameters()` for `anova()` from *lmerTest* models.

# parameters 0.6.0

## Breaking changes

- Alias `model_bootstrap()` was removed, please use `bootstrap_model()`.

- Alias `parameters_bootstrap()` was removed, please use
  `bootstrap_parameters()`.

- Alias `model_simulate()` was removed, please use `simulate_model()`.

- Alias `parameters_simulate()` was removed, please use
  `simulate_parameters()`.

- Alias `parameters_selection()` was removed, please use `select_parameters()`.

- Alias `parameters_reduction()` was removed, please use `reduce_parameters()`.

- Functions `DDR()`, `ICA()` and `cmds()` are no longer exported, as these were
  intended to be used internally by `reduce_parameters()` only.

- `skewness()` and `kurtosis()` always return a data frame.

## New supported models

- Added support for `arima` (*stats*), `bife` (*bife*), `bcplm` and `zcpglm`
  (*cplm*)

## Changes to functions

### model_parameters()

- Improved print-method for `model_parameters.brmsfit()`.

- Improved print-method for `model_parameters.merMod()` when fitting REWB-Models
  (see `demean()`).

- Improved efficiency for `model_parameters()` (for linear mixed models) when
  `df_method = "kenward"`.

- `model_parameters()` gets a `p_adjust`-argument, to adjust p-values for
  multiple comparisons.

- Minor improvements for `cluster_analysis()` when `method = "kmeans"` and
  `force = TRUE` (factors now also work for kmeans-clustering).

### p_value(), ci() and standard_error()

- `p_value_kenward()`, `se_kenward()` etc. now give a warning when model was not
  fitted by REML.

- Added `ci()`, `standard_error()` and `p_value()` for *lavaan* and *blavaan*
  objects.

- Added `standard_error()` for *brmsfit* and *stanreg* objects.

### Other changes

- Run certain tests only locally, to reduce duration of CRAN checks.

- `skewness()`, `kurtosis()` and `smoothness()` get an `iteration` argument, to
  set the numbers of bootstrap replicates for computing standard errors.

- Improved print-method for `factor_analysis()`.

- `demean()` now additionally converts factors with more than 2 levels to
  dummy-variables (binary), to mimic *panelr*-behaviour.

## Bug fixes

- Fixed minor issue with the `print()`-method for `model_parameters.befa()`.

- Fixed issues in `model_parameters()` (for linear mixed models) with wrong
  order of degrees of freedom when `df_method` was different from default.

- Fixed issues in `model_parameters()` (for linear mixed models) with accuracy
  of p-values when `df_method = "kenward`.

- Fixed issues in `model_parameters()` with wrong test statistic for
  *lmerModLmerTest* models.

- Fixed issue in `format_parameters()` (which is used to format output of
  `model_parameters()`) for factors, when variable name was also part of factor
  levels.

- Fixed issue in `degrees_of_freedem()` for *logistf*-models, which
  unintentionally printed the complete model summary.

- Fixed issue in `model_parameters()` for *mlm* models.

- Fixed issue in `random_parameters()` for uncorrelated random effects.