Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/mlr-org/mlr3pipelines
Dataflow Programming for Machine Learning in R
https://github.com/mlr-org/mlr3pipelines
bagging data-science dataflow-programming ensemble-learning machine-learning mlr3 pipelines preprocessing r r-package stacking
Last synced: about 1 month ago
JSON representation
Dataflow Programming for Machine Learning in R
- Host: GitHub
- URL: https://github.com/mlr-org/mlr3pipelines
- Owner: mlr-org
- License: lgpl-3.0
- Created: 2017-10-10T19:01:45.000Z (almost 7 years ago)
- Default Branch: master
- Last Pushed: 2024-05-28T11:26:57.000Z (4 months ago)
- Last Synced: 2024-05-29T02:24:20.541Z (4 months ago)
- Topics: bagging, data-science, dataflow-programming, ensemble-learning, machine-learning, mlr3, pipelines, preprocessing, r, r-package, stacking
- Language: R
- Homepage: https://mlr3pipelines.mlr-org.com/
- Size: 16.2 MB
- Stars: 130
- Watchers: 18
- Forks: 25
- Open Issues: 142
-
Metadata Files:
- Readme: README.Rmd
- License: LICENSE
Awesome Lists containing this project
- jimsghstars - mlr-org/mlr3pipelines - Dataflow Programming for Machine Learning in R (R)
README
---
output: github_document
---
# mlr3pipelines 
Package website: [release](https://mlr3pipelines.mlr-org.com/) | [dev](https://mlr3pipelines.mlr-org.com/dev/)
Dataflow Programming for Machine Learning in R.
[](https://github.com/mlr-org/mlr3pipelines/actions/workflows/r-cmd-check.yml)
[](https://cran.r-project.org/package=mlr3pipelines)
[](https://stackoverflow.com/questions/tagged/mlr3)
[](https://lmmisld-lmu-stats-slds.srv.mwn.de/mlr_invite/)
```{r, include = FALSE}
knitr::opts_chunk$set(
cache = FALSE,
collapse = TRUE,
comment = "#>"
)
set.seed(8008135)
library("paradox")
library("mlr3")
library("mlr3pipelines")
library("mlr3learners")
lgr::get_logger("mlr3")$set_threshold("warn")
```
## What is `mlr3pipelines`?
Watch our "WhyR 2020" Webinar Presentation on Youtube for an introduction! Find the slides [here](https://raw.githubusercontent.com/mlr-org/mlr-outreach/main/2020_whyr/slides.pdf).
[](https://www.youtube.com/watch?v=4r8K3GO5wk4)
**`mlr3pipelines`** is a [dataflow programming](https://en.wikipedia.org/wiki/Dataflow_programming) toolkit for machine learning in R utilising the **[mlr3](https://github.com/mlr-org/mlr3)** package. Machine learning workflows can be written as directed "Graphs" that represent data flows between preprocessing, model fitting, and ensemble learning units in an expressive and intuitive language. Using methods from the **[mlr3tuning](https://github.com/mlr-org/mlr3tuning)** package, it is even possible to simultaneously optimize parameters of multiple processing units.
In principle, *mlr3pipelines* is about defining singular data and model manipulation steps as "PipeOps":
```{r}
pca = po("pca")
filter = po("filter", filter = mlr3filters::flt("variance"), filter.frac = 0.5)
learner_po = po("learner", learner = lrn("classif.rpart"))
```
These pipeops can then be combined together to define machine learning pipelines. These can be wrapped in a `GraphLearner` that behave like any other `Learner` in `mlr3`.
```{r}
graph = pca %>>% filter %>>% learner_po
glrn = GraphLearner$new(graph)
```
This learner can be used for resampling, benchmarking, and even tuning.
```{r}
resample(tsk("iris"), glrn, rsmp("cv"))
```
## Feature Overview
Single computational steps can be represented as so-called **PipeOps**, which can then be connected with directed edges in a **Graph**. The scope of *mlr3pipelines* is still growing; currently supported features are:
* Simple data manipulation and preprocessing operations, e.g. PCA, feature filtering
* Task subsampling for speed and outcome class imbalance handling
* *mlr3* *Learner* operations for prediction and stacking
* Simultaneous path branching (data going both ways)
* Alternative path branching (data going one specific way, controlled by hyperparameters)
* Ensemble methods and aggregation of predictions
## Documentation
A good way to get into `mlr3pipelines` are the following two vignettes:
* [Sequential Pipelines](https://mlr3book.mlr-org.com/chapters/chapter7/sequential_pipelines.html)
* [Non-Sequential Pipelines and Tuning](https://mlr3book.mlr-org.com/chapters/chapter8/non-sequential_pipelines_and_tuning.html)
## Bugs, Questions, Feedback
*mlr3pipelines* is a free and open source software project that encourages participation and feedback. If you have any issues, questions, suggestions or feedback, please do not hesitate to open an "issue" about it on the GitHub page!
In case of problems / bugs, it is often helpful if you provide a "minimum working example" that showcases the behaviour (but don't worry about this if the bug is obvious).
Please understand that the resources of the project are limited: response may sometimes be delayed by a few days, and some feature suggestions may be rejected if they are deemed too tangential to the vision behind the project.
## Citing mlr3pipelines
If you use mlr3pipelines, please cite our [JMLR article](https://jmlr.org/papers/v22/21-0281.html):
```{r echo = FALSE, comment = ""}
toBibtex(citation("mlr3pipelines"))
```
## Similar Projects
A predecessor to this package is the [*mlrCPO*-package](https://github.com/mlr-org/mlrCPO), which works with *mlr* 2.x. Other packages that provide, to varying degree, some preprocessing functionality or machine learning domain specific language, are the *[caret](https://github.com/topepo/caret)* package and the related *[recipes](https://recipes.tidymodels.org/)* project, and the *[dplyr](https://github.com/tidyverse/dplyr)* package.