parameters
parameters copied to clipboard
easystats
:bar_chart: Computation and processing of models' parameters
`include_reference = TRUE` should only add the referent when standard dummy coding is used, but for some reason it also adds a (wrong) 0 when `datawizard::contr.deviation()` is used. ```R contrs
Fixes #962
Note the difference between `model_parameters()` and `standardize_parameters()`: ``` r library(easystats) #> # Attaching packages: easystats 0.7.2 #> ✔ bayestestR 0.13.2 ✔ correlation 0.8.4 #> ✔ datawizard 0.11.0 ✔ effectsize 0.8.8...
For some reason the equivalent tests says Accepted when the low bound becomes "big" (?) ``` r m #> ROPE: [-0.08 0.08] #> #> Parameter | 90% CI | SGPV...
parameters:::.extract_htest_correlation should always return CI info if the htest object provides it
Currently, [`parameters:::.extract_htest_correlation`](https://github.com/easystats/parameters/blob/main/R/methods_htest.R#L200-L229) only returns confidence intervals for `method = "pearson"`, which makes sense since the default `cor.test` only returns CIs for this method. However, I implemented an alternative Spearman correlation...
The current default for mixed models is to use residual dfs, but these are counter conservative. Should we default (when possible) to use Satterthwaite or Kenward-Roger dfs instead?
In one of the recent `bayestestR` PRs (https://github.com/easystats/bayestestR/pull/673 & https://github.com/easystats/bayestestR/pull/672) some of the printing methods have adopted allowing for arbitrary columns in the resulting data frame objects. This was done...
Related to https://github.com/easystats/modelbased/issues/213