The iraceplot package

Maintainers: Leslie Pérez Cáceres, Manuel López-Ibáñez

Creators: Pablo Oñate Marín, Leslie Pérez Cáceres, Manuel López-Ibáñez

Contact: https://groups.google.com/d/forum/irace-package

Introduction

The iraceplot package provides different plots to visualize the data generated by the irace software for automatic algorithm configuration (hyper-parameter optimization).

This package provides visualizations of:

• Parameter configurations using parallel coordinates with parallel_coord().
• Boxplots of configurations performance (training and testing) with boxplot_training() and boxplot_test().
• Sampling distributions with sampling_frequency().
• Overview of all performance data generate by a single run of irace with plot_experiments_matrix().

For more details about these functions, please check the user guide of the package and the documentation of the functions implemented in the package.

The package also provides an HTML report, using report(), summarizing relevant information obtained during an execution of irace.

The aim of this package is to provide support for the analysis of the best parameter settings found, the assessment of the parameter space explored by irace and the overall performance of the configuration process. Such analysis might lead to insights about the role of algorithmic components their interactions, or to improve the configuration process itself.

Keywords: automatic configuration, offline tuning, parameter tuning, parameter visualization, irace.

Requisites

• R (https://www.r-project.org) is required for running irace and to use iraceplot.

User guide

A user guide comes with the package. The following is a quick-start guide. The user guide gives more detailed instructions.

Installing R

The official instructions are available at https://cran.r-project.org/doc/manuals/r-release/R-admin.html. We give below a quick R installation guide that will work in most cases.

GNU/Linux

You should install R from your package manager. On a Debian/Ubuntu system it will be something like:

$ sudo apt-get install r-base

Once R is installed, you can launch R from the Terminal and from the R prompt install the iraceplot package. See instructions below.

OS X

You can install R directly from a CRAN mirror (https://cran.r-project.org/bin/macosx/).

Alternatively, if you use homebrew, you can just do

    $ brew install --cask r

(Using brew install r is not recommended because that will build R from source and you will not be able to use any CRAN binary, possibly resulting in annoying build failures).

Once R is installed, you can launch R from the Terminal (or from your Applications), and from the R prompt install the iraceplot package. See instructions below.

Windows

You can install R from a CRAN mirror (https://cran.r-project.org/bin/windows/). Once R is installed, you can launch the R console and install the iraceplot package from it. See instructions below.

Installing the iraceplot package

Stable version

For installing the stable version from CRAN, launch R or Rstudio and evaluate:

install.packages("iraceplot")

Or you may wish to try the development version from GitHub

install.packages("devtools")
devtools::install_github("auto-optimization/iraceplot")

The above commands will install many packages so they may take a while.

Basic Usage

Load the package in the R console:

library(iraceplot)

You need the log file generated by irace (irace.Rdata or the filename given by the option logFile of irace). Then, generate a general-purpose report with:

report("irace.Rdata")

This should create a filename report.html and open it in your browser (if the browser does not open, try to find the file and open it yourself). The result will look something like this report example.

There is a lot more functionality in this package. Check the documentation and the User Guide to identify the plots most suited to your needs.