I found out about MNE recently after becoming fed up of the headache of
Matlab for EEG/MEG work (I'm a long term Python user). I've read up about
it in the recent NeuroImage paper, and am impressed with the number of
features and like the philosophy!
Sorry if this has been raised before (couldn't find an issue for it on the
Github), but is there plans for Python 3 support? The rest of my python
analysis code is now in Py3, since the majority of major scientific
libraries now support it (numpy, scipy, matplotlib, pandas, scikit-learn
etc.), and sometime in the future it would be good to not have to go back
and forth between Py2/3.
I can help out with porting if there is no major changes required, and
people are interested in maintaining a separate branch.
I found out about MNE recently after becoming fed up of the headache of
Matlab for EEG/MEG work (I'm a long term Python user). I've read up about it
in the recent NeuroImage paper, and am impressed with the number of features
and like the philosophy!
that's great to hear !
Sorry if this has been raised before (couldn't find an issue for it on the
Github), but is there plans for Python 3 support? The rest of my python
analysis code is now in Py3, since the majority of major scientific
libraries now support it (numpy, scipy, matplotlib, pandas, scikit-learn
etc.), and sometime in the future it would be good to not have to go back
and forth between Py2/3.
it would be great to port it to Python 3. It has not been done yet
because nobody asked for it. The core devs still use 2.7.
If you could start to take a stab at it and send us a pull request, it
would be break.
we'll enable python 3 on travis for continuous integration.
the only (thankully weak) dependency that will not work is Mayavi.
thanks for your feedback, this is nice to hear!
Your request is very timely. Indeed, we were thinking about a Python 3
port as one of the potential goals / milestones for the next release
cycle.
As Alex mentioned, Mayavi2 which we use as 3D visualization backend is
a blocker to Python 3 support. This is because, most probably, it will
never be ported to Python 3. However, I'm confident a solution can be
found to this.
I have had great success with pyqtgraph to do interactive 2/3D
visualizations, and it's pure-Python both v. 2 & 3. However, the 3D support
is mainly a nice layer over OpenGL, not the extensively structured pipeline
that's in VTK. Nevertheless, this has worked well for visualizing the
cortical surface and M/EEG forward solutions that we generate in TVB (and
will soon analyze in MNE!) and could share the code if interested.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20131128/46b3abdc/attachment.html