R-Forge Project: Probabilistic Index Models - News

pim development moved to github

jorismeys@users.r-forge.r-project.org (Joris Meys) — Fri, 25 Aug 2017 15:33:04 GMT

In order to centralize the development of our packages, we decided to move everything to github. This SVN repo on R-Forge will be updated with every new release on CRAN. If you want to get the latest dev version, please visit

https://github.com/CenterForStatistics-UGent/pim

Thank you!

pim from version 1.9.xxx with new formula interface

jorismeys@users.r-forge.r-project.org (Joris Meys) — Mon, 22 Feb 2016 08:10:52 GMT

As announced, the development package pimdevel is moved to the main branch and is as of today the official pim package. This package will be available from CRAN soon.

The original pim package (version 1.1.5.6) is renamed to pimold and is only kept for legacy purposes. It contains the pim package in the version used for earlier publications.

Unless you want to try out the example code of old publications, please use the new pim package for model fitting. This is more stable and has a more flexible formula interface.

Cheers
The pim team

renaming pim and pimdevel

jorismeys@users.r-forge.r-project.org (Joris Meys) — Mon, 08 Feb 2016 11:11:00 GMT

On february 15, the current pim package will be renamed pim_old, and the pimdevel package will become the new pim package. It will be kept at 1.9.xxx until stability is confirmed, and will then be pushed to CRAN as 2.0

If your code still depends on the old pim package, please take a look at pimdevel to get acquainted with the new interface.

Cheers
Joris Meys

Big overhaul of the codebase

jorismeys@users.r-forge.r-project.org (Joris Meys) — Tue, 24 Jun 2014 12:38:41 GMT

In order to get the pim package CRAN compliant and to improve the user interface, we decided to overhaul the complete code base. This includes the disappearance of experimental and non-published functionality in the official pim package.

For researchers, the current implementation stays available as pimdevel on R-forge. This package is mainly directed to researchers that want to implement and use their own estimators, and/or need more flexibility in the specification of exotic pim models.

The codebase overhaul will lead eventually to pim 2.0 to be released on CRAN.

Cheers