Skip to content

Releases: choderalab/pymbar

Critical bugfix release

02 Dec 01:24
1d14e80

Choose a tag to compare

In adding support for python 3.8 and scipy 1.0, pymbar 3.0.4 accidentally changed the API for key components to a new experimental API in which dict objects were returned for convenience.

This release restores the previous API behavior in pymbar 3.0.0-3.0.3, but allows the new API to be accessed via the optional keyword argument return_dict=True (via PR #319). Without this keyword, the old API is used.

We will consider making the dict API the default in pymbar 4, but we will plan to add warnings of the API change well in advance of this change.

Apologies to all our users who were affected by this issue!

Python 3.8 and Newer Scipy 1.0 compatibility

14 Nov 19:21
d1c18f3

Choose a tag to compare

This release brings in all the changes from the last months and makes PyMBAR compatible with Python 3.8 and Scipy 1.0.

Fix PyPi Release

06 Sep 18:02

Choose a tag to compare

This release fixes a subtle but important issue caused by missing file when PyPi Tarball was compiled from Python 3, but then downloaded in Python 2

Because this changes the setup.py file, this formal release was cut.

3.0.2 Updated Numpy Support and Conda Forge Version

29 Jun 22:05

Choose a tag to compare

Merge pull request #266 from choderalab/PrepCFRelease

3.0.2 release and bugfix

3.0.1.beta0 Fixes for Numpy 1.12 and Python 3.6

03 Mar 17:31

Choose a tag to compare

Cut a release to bring in Numpy 1.12 changes and Python 3.6 support for deployment.

3.0.0 beta 2

16 Jun 15:58

Choose a tag to compare

3.0.0 beta 2 Pre-release
Pre-release

This release fixes win conda package support.

3.0.0 beta 1

10 Jun 15:14

Choose a tag to compare

3.0.0 beta 1 Pre-release
Pre-release
Merge pull request #189 from mrshirts/master

BAR improvements.

Beta release of conda-packageable pymbar distribution

10 Jun 14:00

Choose a tag to compare

A number of minor improvements (including some API changes) have been made in this version:

  • Docstrings now conform to the NumPy/SciPy convention
  • Documentation has been cleaned up
  • Installation of prerequisites has been automated
  • Pyflakes was used to remove unused dependencies
  • Tests have been improved and automated using nose
  • Automated integration testing implemented with travis-ci
  • License has been changed to LGPL
  • Import and init styles have been made more consistent
  • Source code now more generally follows pep8
  • Test systems have been refactored