-
Notifications
You must be signed in to change notification settings - Fork 8
/
NEWS
71 lines (51 loc) · 2.62 KB
/
NEWS
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
0.2.4
===========
* Adding tests for rboot noise feature.
0.2.3
===========
* There are no user-visible changes. This fixes an error with testing on CRAN based on RNG differences.
0.2.3
===========
BUG FIXES
* `rbootnoise` argument added to residual bootstrap to bypass "system is exactly singular" issues with lme4 models (Credit to Ilmari Tamminen)
* Eliminated unnecessary warnings for unnamed vectors with type = "reb2".
* `bootstrap_pvals()` added mean centering for correction to the calculation
0.2.2
===========
NEW FEATURES
* Additional auxilliary distributions were added for the Wild bootstrap. There are now 6 options, including the standard normal.
* `.refit` argument can be set to FALSE in order to return the only bootstrap responses.
* `bootstrap_pvals()` appends bootstrap p-values to the summary table for the fixed effects
* `combine_pvals()` provides a way to combine the results of `bootstrap_pvals()` for parallel runs.
BUG FIXES
* `plot.lmeresamp()` now works if the replicates are a numeric vector rather than a data frame or tibble.
* bug fixed when `na.action = na.omit`
* fixed issue with transformed variables in `glmer`
0.2.1
===========
* Unarchiving from CRAN
DEPENDENCY CHANGE
* Remove `catchr` dependency to avoid issues on CRAN
BUG FIXES
* message/error/warning summarization in `summary.lmeresamp` has been fixed
* If `var` is omitted from `plot.lmeresamp()` a halfeye plot with all terms is created.
0.2.0
=====
* The case, parametric, and residual bootstraps now suppport `glmerMod` objects.
* The Wild bootstrap is available for `lme` and `lmerMod` objects.
* The CGR bootstrap is now the default "residual" bootstrap algorithm.
* Objects returned by the `bootstrap()` call are now of class `lmeresamp`.
* `lmeresamp` objects have a new structure, including a new `stats` dataframe (contains the observed value, bootstrap mean, standard error, and bias of each LME model parameter).
* New generic `print()` function that is compatible with `lmeresamp` objects
* New generic `confint()` function that is compatible with `lmeresamp` objects (the possible confidence intervals include: basic, normal, percentile, or all)
* A package vignette is now available
* Vignette outlines how to perform parallelization in `bootstrap()` using the `doParallel` and `foreach` packages
* New `combine()` function that combines processes split for parallelization for unified output
0.1.1
=====
* Unarchiving from CRAN - back to active development
* Updating for use with the new version of dplyr (>= 0.8.0)
* Bug fixed for `case_bootstrap.lme ` so that `.cases.resamp` can be found
Version 0.1.0
=============
* Initial release, enjoy!