stacks
stacks copied to clipboard
An R package for tidy stacked ensemble modeling
When making patches to fix runs on `general-meta`, I noticed that multisession workers required that needed parsnip extension packages be loaded via `options.future` inside of `fit_members()`: https://github.com/tidymodels/stacks/blob/f2e33419b5712ae7df3ca4d689c652188dce4b59/R/fit_members.R#L157 I wonder if...
After finalizing a model_stack with fit_members(), I get a large object (~13.5 GB) that is inconvenient to use in my specific research production setting. Is there a way to trim...
Hi @simonpcouch, I love the `{stacks}` package and over the last several months have thought about it quite a bit and whether there is room to broaden the API to...
`use_standalone("r-lib/rlang", "types-check")` preferred over home-grown argument checkers.
Some explanation of the internal data structures and how different senses of train/test splits are utilized in the package—finding that it would be nice to point to something like this....
tips/tricks to stack models quickly for preliminary results—racing / choosing grids, choosing model specs, possible results of #119
Prepare for release: * [x] `git pull` * [x] Check [current CRAN check results](https://cran.rstudio.org/web/checks/check_results_stacks.html) * [x] [Polish NEWS](https://style.tidyverse.org/news.html#news-release) * [x] `urlchecker::url_check()` * [x] `devtools::build_readme()` * [x] `devtools::check(remote = TRUE, manual...
In many places in stacks, the package name is written "`stacks`" rather than "stacks" (i.e. there are backticks before and after). Per [tidy style](https://style.tidyverse.org/index.html), package names don't need backticks. Here's...
Simon hi - I wanted to label `autoplot("weights")` with the names of the engines. Use of `stacks:::top_coefs()` was very handy for getting the names! I was curious though why the...
Hi - Stacks is proving very effective at improving model performance for me (using the output from `finetune::tune_sim_anneal()`). Thank you! I am occasionally running into this problem though stacking classification...