[solved] Inline script and Python code give different results

ghostrider

Shuffle seem so work differently in OpenSesame inline script than in python.

In python (PyCharm) the following code shuffles the list with no repeats, so that when I slice the list there are no duplicates.

import random

picList = ["achievement.jpg", "admired.jpg", "alert.jpg", "cope.jpg", "energy.jpg", 
"friendships.jpg","good_decision.jpg", "good_news.jpg", "good_times.jpg", "healthy.jpg", 
"hope.jpg", "liked.jpg", "praised.jpg", "problem_solve.jpg", "success.jpg"]


random.shuffle(picList)

list2 = picList[5:]

In my Opensesame the same code gives repeats on the shuffle. Any ideas? What am I doing wrong?

sebastiaan

Hi,

There's no difference between (the same version of) Python in and outside of OpenSesame. Essentially, OpenSesame itself is simply a Python library, and an inline_script item uses the official Python interpreter to execute its script.

But, as you say, random.shuffle() doesn't give repeats. Unless, of course, there were repeats in the list to begin with. Could that be it?

Cheers,
Sebastiaan

ghostrider

I am not sure how that can be? The list is there, and there are no repeats in it.

ghostrider

This is the whole code, part of a loop sequence (30 repeats) and followed in the sequence by a sketchpad set to draw "[pic]" , and a logger.

import random


picList = ["achievement.jpg", "admired.jpg", "alert.jpg", "cope.jpg", "energy.jpg", 
"friendships.jpg","good_decision.jpg", "good_news.jpg", "good_times.jpg", "healthy.jpg", 
"hope.jpg", "liked.jpg", "praised.jpg", "problem_solve.jpg", "success.jpg"]


random.shuffle(picList) #shuffle picList

list2 = picList[5:] #list2 from picList element 5 up (10 items)


list3= list2[:5] #list3 from list2, first 5 elements

list4 = list2[5:] #list4 from list2 second five elements

cue = (list3 *5) + list4 #cue list3 repeated 5 times (25 elements) + list 4 (5 elements) 

pic= cue[0]

exp.set("pic", pic)

There must be something wrong, but I do not know what.

ghostrider

Hoping to get a final cue list of thirty elements 5 x 5 + 5 x1 but getting various results
seen here. (I have sorted them to make checking easier).

achievement.jpg
achievement.jpg
admired.jpg
alert.jpg
cope.jpg
cope.jpg
friendships.jpg
friendships.jpg
friendships.jpg
good_decision.jpg
good_news.jpg
good_news.jpg
good_news.jpg
good_times.jpg
healthy.jpg
healthy.jpg
healthy.jpg
healthy.jpg
healthy.jpg
hope.jpg
hope.jpg
hope.jpg
liked.jpg
liked.jpg
praised.jpg
praised.jpg
praised.jpg
praised.jpg
problem_solve.jpg
problem_solve.jpg

sebastiaan

It's difficult to say without seeing the rest of the experiment, but what may be going on here is that you are reshuffling the list on every trial, instead of only shuffling once and then walking through the shuffled one item at a time.

Basically, if you reshuffle your list in a separate inline_script that is executed before a block of trials, and then pop() (=select last and remove from list) an item from the list in an inline_script that is part of the trial, you will not encounter any repeats. However, if you reshuffle the list also on every trial, you will sometimes get repeats. And this appears to be what you're doing, because you're reshuffling and selecting in the same script, which I assume is executed on every trial. Does that make sense?

ghostrider

I am sure you are right Sebastiaan, but I am not sure how to correct it. I have provided a link to the experiment, I would be so grateful if you could take a look and confirm your theory.

I will keep trying to figure it out.

https://www.dropbox.com/s/xjttoewxkush7eb/mentalsimulation.opensesame.tar.gz

ghostrider

this is a cut down version with just the loop , that I use to test

https://www.dropbox.com/s/nql4yubsek0p7eo/mentalsimulation2.opensesame.tar.gz

sebastiaan

Hi,

Yes, so it's basically what I suspected: You're creating a new shuffled list on every trial, instead of once at the start of the block.

Let's just focus on the simple case, in which you want to loop through picList in a random order. First, before the block loop (called Initial_List in your experiment), create a randomly ordered list of pictures, with a simple inline_script.

import random
# Create a list of pictures ...
picList = ["achievement.jpg", "admired.jpg", "alert.jpg", "cope.jpg",
    "energy.jpg", "friendships.jpg","good_decision.jpg", "good_news.jpg",
    "good_times.jpg", "healthy.jpg", "hope.jpg", "liked.jpg", "praised.jpg",
    "problem_solve.jpg", "success.jpg"]
# ... and shuffle the order
random.shuffle(picList)

Then, in a different inline_script at the start of the trial sequence (called sequence in your experiment), retrieve one item from picList. The pop() function, which returns the last item from a list and removes it from the list, is convenient here.

# Take the last picture from the list ...
pic = picList.pop()
# .. and set it as the experiment variable `pic`.
exp.set("pic", pic)

So that's basically it. It appears that you have something more complicated in mind, because you are using multiple lists that contain different slices of the picList. But this at least shows how you can avoid reshuffling your list again on every trial.

Cheers!
Sebastiaan

ghostrider

Thank you so much, and sorry for being a pain. You have put me on the right track. and now I have been able to sort it. The practice run and the main run need 3 scripts each, two before the loop and one during the sequence

script 1

import random


picList = ["achievement.jpg", "admired.jpg", "alert.jpg", "cope.jpg", "energy.jpg", 
"friendships.jpg","good_decision.jpg", "good_news.jpg", "good_times.jpg", "healthy.jpg", 
"hope.jpg", "liked.jpg", "praised.jpg", "problem_solve.jpg", "success.jpg"]


random.shuffle(picList)

script 2

list2 = picList[5:]


list3= list2[:5]

list4 = list2[5:]

cue = (list3 *5) + list4

picList = cue

random.shuffle(picList)

script 3

pic = picList.pop()

exp.set("pic", pic)

The scripts need to be named differently and have different variable names for the practice run, because the list was much shorter (5 images).

Thank you once again :)