README.qmd

---
format: gfm
default-image-extension: ""
editor_options: 
  chunk_output_type: console
---

# mllrnrs

<!-- badges: start -->
```{r}
#| echo: false
#| message: false
#| results: asis

pkg <- desc::desc_get_field("Package")
cat_var <- paste(
  badger::badge_lifecycle(),
  badger::badge_cran_release(pkg = pkg),
  gsub("summary", "worst", badger::badge_cran_checks(pkg = pkg)),
  badger::badge_cran_download(pkg = pkg, type = "grand-total", color = "blue"),
  badger::badge_cran_download(pkg = pkg, type = "last-month", color = "blue"),
  gsub("netlify\\.com", "netlify.app", badger::badge_dependencies(pkg = pkg)),
  badger::badge_github_actions(action = utils::URLencode("R CMD Check via {tic}")),
  badger::badge_github_actions(action = "lint"),
  badger::badge_github_actions(action = "test-coverage"),
  badger::badge_codecov(ref = desc::desc_get_urls()),
  sep = "\n"
)
cat_var |> cat()
```

The `mllrnrs` R package ships with additional ML learners for [`mlexperiments`](https://github.com/kapsner/mlexperiments). 

Currently implemented learners are:

| Name | Based on | Description / Tasks |
| ---- | -------- | ------------------- |
| LearnerGlmnet | `glmnet::glmnet` | General interface to `glmnet` (examples available for families [`binomial`](tests/testthat/test-glmnet_binary.R) , [`multinomial`](tests/testthat/test-glmnet_multiclass.R), and [`regression`](tests/testthat/test-glmnet_regression.R) |
| LearnerLightgbm | `lightgbm::lgb.train` | General interface to `lightgbm` (examples available for objectives [`binary`](tests/testthat/test-lightgbm_binary.R) , [`multiclass`](tests/testthat/test-lightgbm_multiclass.R), and [`regression`](tests/testthat/test-lightgbm_regression.R) |
| LearnerRanger | `ranger::ranger` | General interface to `ranger` (examples available for tasks [`binary`](tests/testthat/test-ranger_binary.R) , [`multiclass`](tests/testthat/test-ranger_multiclass.R), and [`regression`](tests/testthat/test-ranger_regression.R) |
| LearnerXgboost | `xgboost::xgb.train` | General interface to `xgboost` (examples available for objectives [`binary:logistic`](tests/testthat/test-xgboost_binary.R) , [`multi:softprob`](tests/testthat/test-xgboost_multiclass.R), and [`reg:squarederror`](tests/testthat/test-xgboost_regression.R) |

For a short introduction on how to use the learners together with the `mlexperiments` R package, please visit the [wiki page](https://github.com/kapsner/mllrnrs/wiki).

Some learner for survival tasks are implemented in the [`mlsurvlrnrs`](https://github.com/kapsner/mlsurvlrnrs) R package.

## Installation

`mllrnrs` can be installed directly from CRAN:

```{r}
#| eval: false
install.packages("mllrnrs")
```

To install the development version, run

```{r}
#| eval: false
install.packages("remotes")
remotes::install_github("kapsner/mllrnrs")
```