I created several psychopy coding files consisting of python language each forming one condition of an experiment. I want that the python files are randomly chosen and each of them is used once. I thought about making a list containing the names of the files in an additional psychopy file, but I am stuck and I am not sure whether this is the right way to go to solve this issue. Your advice would be very much appreciated.
If you want a random element from a list use random.choice where the list would be the names of the files. You can delete elements from the list as you go.
Or in the shell use ls to list files then shuf to reorder. It would look like
for F in $(ls | shuf); do bash $F ; done
It also seems to me a robust solution. Take a look into "listdir" from "os" library to automatically retrieve all files in a path:
https://docs.python.org/2/library/os.html#os.listdir
You can also use "subprocess" library (class Popen) to launch those python scripts as external processes:
https://docs.python.org/2/library/subprocess.html#subprocess.Popen
Probably a more gracious way to do this is to run those python scripts after importing them as modules using "__import__" function:
https://docs.python.org/2/library/functions.html#__import__
, and therefore building a "run" function inside those scripts (if you don't already have one). So my suggestion would be:
import os
import random
files = os.listdir(os.getcwd()) # Getting all files inside working directory folder.
files.remove('thisscriptname.py') # Removing this script name from the list.
choice = random.choice(files) # Chose random file.
choice_module = __import__(choice) # Import file as module.
choice_module.run() # Run function that runs the code in the script.
Hope it helps.
PS: I didn't test the code.
PS2: Remember to save the script into the folder where you have the other files.
Thank you very much for your advices and confirming that this path is possible to solve my issue.
I suggest using the new pathlib.Path (in python 3.4, but also available for python 2.7 as an installable package) to simplify getting files in folders and subfolders.
For example, to get a list of all .py files in current folder an subfolders:
file_list = list(Path('**/*.py'))
Once you have the list, you can directly shuffle it with "random.shuffle".
https://docs.python.org/3/library/pathlib.html
https://docs.python.org/3.4/library/random.html#random.shuffle
Thank you very much for your reply
I will see what the best option for me is.
In case this topic is still of interest: I found another solution to sequentially start several psychopy scripts (if you want to: in a random sequence). It is described in this discussion:
https://groups.google.com/forum/#!topic/psychopy-users/3V8YUAwsdIs
Basically, you import sys and core:
import sys
from psychopy import core
then you start your script with the command:
core.shellCall([sys.executable, "your_script.py"])
Remember that you should set the working directory to the path of the script before that using the os.chdir() command.
I like this solution because it allows you to run your scripts as they are (you don't have to turn them into functions or modules). And errors are avoided because they are started and terminated. So you have less problems with double assignments of commonly used variable names that appear in almost every psychopy script (for example "win").
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages
- Python