Uploading your experiment to Pack&Go

Here are a few things you should keep in mind when preparing your MATLAB experiment for Pack&Go:

• If your experiment contains subfolders, remember to replace any local file paths in your code with relative paths. Do not hardcode your file paths. For more details, see Best Practices for Designing and Implementing Pack&Go Studies

• To retrieve the full path to the Pack&Go working directory, use the pwd function, not the mfilename one. The latter will create an error.

• Pack&Go does not support the MATLAB addpath function.

• Save all result files in the main experiment folder. Do not create result subfolders.

• Pack&Go runs on a remote server with an optimized Linux X64 operating system. If you are using third-party MEX files in your experiment, make sure you specifically include Linux X64 MEX files.

Pack&Go will automatically point to the main experiment folder; when setting up relative paths to subfolders, you can start indexing from the first subfolder and use, for example, the MATLAB filesep function to add platform-specific file separators when needed.

Here are a few things you should keep in mind when preparing your Python/PsychoPy experiment for Pack&Go:

• If your experiment contains subfolders, remember to replace any local file paths in your code with relative paths. Do not hardcode your file paths. For more details, see Best Practices for Designing and Implementing Pack&Go Studies

• If you are using keyboard input in your study, use the Keyboard class from PsychoPy, not the event class. E.g.
  Bad:

  PY

  from psychopy import event
  
  # Wait for spacebar press to start
  event.waitKeys(keyList=['space'])

  Good:

  CODE

  from psychopy.hardware import keyboard   
  kb = keyboard.Keyboard()
  
  # Wait for spacebar press to exit
  kb.waitKeys(keyList=['space'])

• If you are playing audio in your PsychoPy experiment, you must initialize the audio by including the following lines at the start of your code:

  PY

  from psychopy import prefs
  prefs.hardware['audioLib'] = ['pygame']
  from psychopy import sound

• Our service uses the X Window System and therefore you will need to add the following at the beginning of the experiment script for optimal timing:

  PY

  import ctypes
  xlib = ctypes.cdll.LoadLibrary("libX11.so")
  xlib.XInitThreads()

• If your experiment uses any libraries besides Python’s standard library and PsychoPy (e.g. numpy and matplotlib), you must also include in the zipped folder, a requirements text file (named “requirements.txt”) which specifies which packages will need to be included on the server to run your experiment. This file should be in the root of the zipped folder, not a subfolder. For example, the following in a text file will tell us to include the version of ‘matplotlib’ 3.7.3 and ‘numpy’ 1.24.4:

  CODE

  matplotlib==3.7.3
  numpy==1.24.4

• For a general explanation about the requirements.txt file and how to create it see this resource: https://learnpython.com/blog/python-requirements-file/

• If you’re not sure which versions exist for a module, you can search for it here: https://pypi.org/

• Python IDE users can try the following to automatically generate their requirements.txt:

  PY

  pip freeze > requirements.txt

  This will generate a list of ALL dependencies currently installed. We recommend reviewing this .txt and eliminating any packages not used in your experiment, in order to optimize performance.