I recently started using psychopy and want to use it without the standalone version but rather import it as a library.
I am using the Anaconda distribution and have already managed to install a new environment following the instructions given here
The problem is that a dependency (pyglet) does not seem to work on an mac OSX (10.11.3) with 64-bit python:
Exception: pyglet is not compatible with 64-bit Python for versions of Mac OS X prior to 10.6.
On the psychopy webpage it also says, that 32bit python is recommended. I however used the Anaconda3 64bit downloader before I knew I would be needing the 32bit version in the future. Now my question is: Can I set up a new environment that uses 32bit and run that in parallel to my 64bit environments? If yes, how?
If no, do I have to delete my Anaconda and get the 32bit version? What would be possible alternatives?
A similar question has been posted here
Thank you all! S.
After searching through the Internet for the rest of the day, I found a solution that works for me. Hopefully this will help other people as well:
According to this forum post the problem with pyglet does not lie in the 64bit but rather in a broken version check of mac OSX. Remember, the exception was:
Exception: pyglet is not compatible with 64-bit Python for versions of Mac OS X prior to 10.6.
... so python and pyglet somehow thought I wanted to run the script on a Mac OSX <10.
Anaconda-based Solution (if you followed this initial installation):
Within the terminal:
source activate psychopyenv # move to your anaconda psychopy environment
conda remove --name psychopyenv pyglet1.2 # delete the current version of pyglet
conda install -c jcarlin pyglet # install pyglet 1.2.4 from a different channel
Afterwards, I was able to run my psychopy scripts in the spyder editor. One strange thing remains: I have to restart the IPython kernel after running the script. Well ... that's for another day to solve.
UPDATE 03.03.2016:
including core.quit() at the end of the script is the key if you do not want to restart the IPython kernel after running a script. Should have noticed earlier ...