HDDM 0.4 Release candidate

[Thomas Wiecki]

Hi all,

We are nearing the final release of HDDM 0.4 with many new features we
are very excited about (see below). Please help us by testing the new
version and reporting any issues to this thread. Your models should
continue to work as before.

You can update to the most recent version (if you have HDDM already
installed by running):

pip install -U --no-deps kabuki
pip install -U --no-deps hddm

If you haven't installed HDDM, see the installation instructions:
http://ski.clps.brown.edu/hddm_docs/index.html#installing

=============
Release Notes
=============

HDDM 0.4
========

License
-----------

HDDM 0.4 is now distributed under the simplified BSD license (see the
LICENSE file) instead of GPLv3.

New features
------------------

* Handling of outliers via mixture model.
http://ski.clps.brown.edu/hddm_docs/howto.html#deal-with-outliers
* New model HDDMRegression to allow estimation of trial-by-trial
regressions with a covariate.
http://ski.clps.brown.edu/hddm_docs/howto.html#estimate-a-regression-model
* New model HDDMStimulusCoding.
http://ski.clps.brown.edu/hddm_docs/howto.html#code-subject-responses
* New model HLBA -- a hierarchical Linear Ballistic Accumulator model
(hddm.HLBA).
* Posterior predictive quantile plots (see model.plot_posterior_quantiles()).

Bugfixes
------------

* model.load_db() is working again.

Thomas & Imri

[DunovanK]

Awesome, congrats! I am in the process of updating right now and have fresh data so today should be a pretty good day.  I'll let you know how things go.

Best,

Kyle

[Thomas Wiecki]

FYI: I have done some more work cleaning up the code for the 0.4 release and to be sure I just put up a 0.4RC2 which will become 0.4 next week if no one reports any problems. Please test!

Thomas

[Thomas Wiecki]

And that's why there are release candidates. Some install problems have been fixed with RC3 so please try again if it was failing before:

pip install -U --no-deps kabuki
pip install -U --no-deps hddm