Repository files navigation

pmlblite is an R interface to thePenn Machine Learning Benchmarks (PMLB) data repository

If you like pmlblite, give it a star, or fork it and contribute!

Check available data sets and download data from the PMLB repository:

library(pmlblite)# Data set namesclassification_dataset_namesregression_dataset_names# Data set summariesclassification_summaryregression_summary# Download features and labels for iris data set in single data frameiris<- fetch_data('iris')iris# Download features and labels for iris data set in separate data structuresiris<- fetch_data('iris',return_X_y=TRUE)iris$x# data frameiris$y# vector

Should work with any recent version of R.

install.packages("devtools")# Install devtools package if necessarylibrary(devtools)devtools::install_github("makeyourownmaker/pmlblite")

The PMLB repository contains a curated collection of data sets for evaluating and comparing machine learning algorithms.These data sets cover a range of applications, and include binary/multi-class classification problems and regression problems,as well as combinations of categorical, ordinal, and continuous features. There are approximately 290 data sets included in the PMLB repositoryand there are no missing values in these data sets.

All binary and multiclass classification data sets are in theclassification category, and all regression data sets are in theregression category.

All data sets are stored in a common format:

• First row is the column names
• Each following row corresponds to an individual observation
• The target column is namedtarget
• All columns are tab (\t) separated
• All files are compressed withgzip to conserve space

The python interface to thePMLB repository is great but can't be used from the R language.

This R library includes summaries of the classification and regression data sets but doesnotinclude any of the PMLB data sets. The data sets can be downloaded using thefetch_data function whichis similar to the corresponding PMLB python function.

Further info:

?fetch_data?regression_summary?classification_summary

If you use PMLB in a scientific publication, please consider citing the following paper:

Randal S. Olson, William La Cava, Patryk Orzechowski, Ryan J. Urbanowicz, and Jason H. Moore (2017).PMLB: a large benchmark suite for machine learning evaluation and comparison.BioData Mining 10, page 36.

I have no affiliation with the authors of PMLB or the University of Pennsylvania.

• Add tests
• Submit library to CRAN

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

Integration of other data repositories are particularly welcome.

• Penn Machine Learning Benchmarks
• OpenMLApproximately 2,500 datasets - available for download usingR module
• UC Irvine Machine Learning Repository
• mlbench: Machine Learning Benchmark Problems
• Rdatasets: An archive of datasets distributed with R
• datasets.load: Visual interface for loading datasets in RStudio from all installed (unloaded) packages
• stackoverflow: How do I get a list of built-in data sets in R?

GPL-2