irace 3.4 released

42 views
Skip to first unread message

Manuel López-Ibáñez

unread,
Mar 30, 2020, 10:05:09 AM3/30/20
to The irace package: Iterated Racing for Automatic Configuration
Dear colleagues,

Version 3.4 of irace is now available from CRAN (https://CRAN.R-project.org/package=irace).

In addition, the development version of irace is now available in github: https://github.com/MLopez-Ibanez/irace

Special thanks to Nguyen Dang (https://sites.google.com/site/nguyenttdangxyz/) for contributing the irace2pyimp tool.

The NEWS section is:

 * `irace2pyimp` function and executable (`irace2pyimp.exe` on Windows) to
   convert .Rdata files generated by irace to the input files required by the
   parameter importance analysis tool PyImp
   (https://github.com/automl/ParameterImportance).
                                             (Nguyen Dang, Manuel López-Ibáñez)

 * Initial configurations may also be provided directly in R using
   `scenario$initConfigurations`
                                                   (Manuel López-Ibáñez)
 
 * Rdata files are saved in version 2 to keep compatibility with older R
   versions.
                                                   (Manuel López-Ibáñez)

 * Fix invalid assert with ordered parameters:
   "value >= 1L && value <= length(possibleValues) is not TRUE"
                                                   (Leslie Pérez Cáceres)
                                               
 * The `irace` executable (`irace.exe` on Windows) is a compiled binary instead
   of a script. On Windows, `irace.exe` replaces `irace.bat`
                                                   (Manuel López-Ibáñez)

 * inst/examples/Spear contains the Spear (SAT solver) configuration scenario.
                                                   (Manuel López-Ibáñez)

 * Fixed bug when reporting minimum maxTime required.
                                           (Reported by Luciana Salete Buriol,
                                            fixed by Manuel López-Ibáñez)

 * Fixed bug detected by assert:
   `all(apply(!is.na(elite.data$experiments), 1, any))` is not TRUE
                     (Reported by Maxim Buzdalov, fixed by Manuel López-Ibáñez)

Best wishes,

    Manuel López-Ibáñez.

Reply all
Reply to author
Forward
0 new messages