NIPY is not only a module for neuroimaing analysis but an umbrella for other Python neuroimaging-related projects -- see https://github.com/nipy and http://www.nipy.org for more information about their releases.
'Close gh-' statements refer to GitHub issues that are available at:
http://github.com/nipy/nipy/issues
The full VCS changelog is available here:
http://github.com/nipy/nipy/commits/master
Abbreviated authors are:
- MB - Matthew Brett
- BT - Bertrand Thirion
- AR - Alexis Roche
- GV - Gaël Varoquaux
- YH - Yarik Halchenko
- 0.3.0 (Saturday 2 February 2013)
- Bugfix, refactoring and compatibility release
- Addition of EM algorithm for mixed effects analysis (BT)
- New high-level GLM class interface (BT)
- nipy diagnostic scripts save PCA and tsdifana vectors to npz file
- Python 3 compatibility through 3.3 (MB)
- Fixes for compatibility with upcoming Numpy 1.7
- Fixes to background and axis specification in visualization tools (GV, BT)
- Fixes and tests for installed nipy scripts (MB)
- Fix to optimization parameters for Realign4D - thanks to bpinsard
- Fix 0 in affine diagonal for TR=0 in affines by default (MB)
- Allow saving of nipy images loaded from nifti files that lack explicit affine (MB)
- Allow slice_order=None to FmriRealign4D when not doing time interpolation (AR); check for valid slice order specification (YH)
- Refactoring of quantile routine to move code out of C library (AR)
- Fix bug in resampling of unsigned int images (AR)
- Custom doctest machinery to work round differences of dtype repr on different platforms, and to skip doctests with optional dependencies (MB)
- Script to run examples for testing (MB)
- Fix for accidental integer division of frametimes in design matrix generation (Fabian Pedregosa)
- Various other fixes and refactorings with thanks from (AR, BT, MB, YH, Yannick Schwartz, Virgile Fritsch)
- 0.2.0 (Sunday 22 July 2012)
- The first ever official release
- > 30 contributors
- > 6 years in development
- 192 issues closed on github
- The first ever official release