Closed simonpcouch closed 8 months ago
Looks like the couple additional skips affected RNG in survival/finetune snaps:
══ Failed ══════════════════════════════════════════════════════════════════════
── 1. Failure ('test-survival-tune_race_anova.R:1007:3'): race tuning (anova) su
Snapshot of code has changed:
old[5:12] vs new[5:12]
# A tibble: 5 x 6
cost_complexity .metric .eval_time mean n std_err
<dbl> <chr> <dbl> <dbl> <int> <dbl>
- 1 0.0000000001 brier_survival 1 0.177 30 0.00707
+ 1 0.0841 brier_survival 1 0.201 30 0.00415
- 2 0.0841 brier_survival 1 0.201 30 0.00415
+ 2 0.0891 brier_survival 1 0.201 30 0.00415
https://github.com/tidymodels/extratests/pull/194 made the opposite change, and sounds like we've seen this wonkiness when testing locally.
Bundled those snap changes into this PR so that we don't continue getting notifs after this is merged, though unrelated to these changes. Will go ahead and merge the accompanying parsnip PR and then this one.
Closes #209, related to tidymodels/parsnip#1078. :)