psychopy.org | Reference | Downloads | Github

Fatal Python error: (pygame parachute) Segmentation Fault)

Mac OS El Capitain
PsychoPy 1.84
Standalone

I am trying to create an fMRI experiment in which the subjects see a tweet about a certain company, then a questionnaire in which they respond either by 1, 2, 3, or 4 (because these are the numbers linked to the fMRI compatible controller). I have 6 different blocks, each of which have 16 different tweets within them (there are 16 unique tweets for each of the 6 companies). There are also 16 questions they respond to as well, however these questions are the same for each block (each block is a company). so the idea is this:

block1(fixation cross (3s) --> tweet (4s) --> question and response (8s) --> loop until all 16 tweets and questions have been seen) ---------> block2 is the identical idea the only difference being the 16 unique tweets to it. then block 3, 4, etc.

So I am using the builder view because I am fairly new to programming, and it has been working fine, however i randomly receive this error message during block1.

python[5310:423012] ApplePersistenceIgnoreState: Existing state will not be touched. New state will be written to /var/folders/r_/jpvs_nts6kgflw5txp58fk7w0000gn/T/org.psychopy.PsychoPy2.savedState
/Applications/PsychoPy2.app/Contents/Resources/lib/python2.7/openpyxl/reader/worksheet.py:322: UserWarning: Unknown extension is not supported and will be removed
warn(msg)
Fatal Python error: (pygame parachute) Segmentation Fault

Can anyone give me some insight as to how to fix this? Thanks a lot.

1 Like

I have python 3.4 also I forgot to mention.

Should I try an use the older versions of psychopy and python?

Your Python 3.4 installation is irrelevant. Standalone PsychoPy uses its own (Python 2.7)
I also don’t _think_that the problem is caused by using too new a version of PsychoPy.

The issue seems to be with your excel file not loading. Could you provide that so we can see if something’s wrong with it?

Hey Jon,

Thanks a lot for the response. I think you should be able to view this link: https://drive.google.com/open?id=0B_1Ia-HAhPx1WF9KNUFvZjZHSGs

The problem is that the program works fine for a while then crashes giving the error.

You’ve got an excel file with emojis in. Is it possible that it just crashes when it gets to those? I’ve never tried rendering emojis before and not sure how they work in standard fonts

Hey Jon, I will try this shortly and get back to you, I hope this is the issue! Thanks a lot.

Chris

I just tested it without the emojis and it seems to go a lot farther into the lock, however it still crashes. It has to be something within one of the excel cells right?

That's my guess, yes. So maybe you need to run your study in a "sequential" loop rather than "random" for now to work out what trial/row is causing the problem.

Hey Jon, thanks a lot for helping me diagnose the problem, I think it was the emojis because when I did sequential all 16 cells ran through. I have another question I was wondering if you could help me with. I want there to be randomness for each block, however the program takes too long because it can show some trials more than once. Is there an easy way to make it so the program only shows each tweet once randomly?

Please indicate the answer that solved this question (there’s a check box somewhere), and then start a new thread for a new question. This helps keep things more organized and allows other people to find the most relevant things.

Sorry about that I cannot seem to find the solved option.

Yeah, a little hard to find. Look for three dots below the post that solved your question. Click the three dots. Then click the check mark inside a box. Thanks!

Found it! Thanks a bunch love this community.

1 Like

@jon I have the identical error message, but I am already running it sequentially with no emojis and the error does not consistently occur at the same spot. I have attached the excel sheet I am using for my loop. Hoping for insight on how to fix this? Thank you! Trial_Robot.xlsx (53.5 KB)

Is it exactly the same error, with a line like this: [quote=“kmiecchr, post:1, topic:972”]
/Applications/PsychoPy2.app/Contents/Resources/lib/python2.7/openpyxl/reader/worksheet.py:32
[/quote]

Or is it just the final line (Pygame parachute) that is the same?

Not exactly the same. Here is the entire message:

python[71571:2384303] ApplePersistenceIgnoreState: Existing state will not be touched.
New state will be written to /var/folders/hp/6s69rw7x5c5cvj0bq3k3n7x0000gn/T/org.psychopy.PsychoPy2.savedState
Fatal Python error: (pygame parachute) Segmentation Fault

OK, I doubt that’s anything to do with loading your conditions file at all. You’d have to tell us more about your study and whether there’s anything unusual about it. Does this happen after many trials? What sort of stimuli do you have. Are there pieces of the task that you can remove to get rid of the problem.

Basically, we’ve got nothing to go on from just that so you’ll need to narrow down the problem a bit for us.

The error occurs at different lengths into the experiment, but usually around 10 trials in.

On each trial there is a stimulus (.png) that lasts for 2 seconds. The stimulus is a different picture every trial (36 different images). Then there are 8 videos (.mp4) that play showing other agents responses, four of the videos are different robot hands on computer mice and four are an identical video showing the equivalent mouse movement on a computer monitor. Essentially creating 4 pairs of videos, each a video with a hand and its accompanying mouse movements. There are not separate videos for every trial, but a repetition of 4 different hand videos for each agent that shows them answering one of four possible options. After watching the videos choices then the participant is able to choose their answer using the mouse.

The videos and/or the images are the only things I can think of that would produce the error. The rest of the information loading from the excel sheet are just strings.

So try to isolate the problem. e.g. Make a copy of your experiment, and replace the video component with something else (e.g. a text stimulus which displays for the same amount of time).

Does the problem go away? If not, keep the videos in but replace the image stimuli.

If it does go away, keep the video stimulus but make it show a constant video rather than one which changes on each trial. Find out if the problem is due to a particular video or just to playing a video a certain number of times. And so on and so on…