psychopy.org | Reference | Downloads | Github

Is there a need to "compile" PsychoPy scripts?

I am learning to use Psychopy because I want to program an eyetracking experiment. I have to use an Eyelink 1000 plus, so I have read the documentation created by SR Research. This docummentation says that it is not recommended to collect real data without first deploying (compilling) the experimental task because of delaying and timing issues.

I would like to know if it is OK to run a Psychopy experiment with the Run green button in order to collect data (not for debugging purposes). I am asking this because I can’t run the .py scripts generated by Psychopy using a Terminal (Mac) nor a Console (Windows), but the experimental tasks runs perfectly in Psychopy Standalone (Windows and Mac).

Regards,

Armel

It would be useful to see the actual quote from the SR documentation, because this doesn’t really apply to Python programs, which aren’t really compiled at all (one of the advantages of using the Python language).

You should avoid running actual scripts in another Python IDE (such as PyCharm). They can be useful in providing a debugger during development, but that extra layer might slow things down in production. But you are perfectly able to run scripts from PsychoPy using the big green button (whether in Builder or Coder views). The PsychoPy application was entirely designed around that being a good (and even preferred) way to run your scripts.

Thank you very much, Michael.

Actually, SR Research documentation says “Deploy”, not “compile” (my mistake). Deploying an experiment designed in Experiment Builder (a software for creating experiments) is required in order to garantee that all the computer resources are used for running the experiment. For this reason, I concluded that in other software (like Psychopy) this “deploying requirement” would be the same. But based in your response, I see that I can run experiments directly from Psychopy Standalone interface.