psychopy | For running psychology and neuroscience experiments | Machine Learning library

Solved (kind of!).

For UI reasons, I stop keys typed during the experiment from being echoed to the screen (otherwise the screen gets full of random junk). I do this by sending stty -echo to the parent shell.

But when exiting from the program (for any reason) I need to undo this by sending stty echo. So I set up an exit handler:

From the error message I believe the sound file you've used here has a sampling rate of 32000 Hz, but you have another sound in your project sampled at 48000 Hz. Switching between sample rates is called resampling but resampling takes time. PsychoPy is designed to provide low-latency playback of sounds but to do this it will not perform any resampling for you. As a result, you should convert all your sounds to use the same sampling rate before the study runs using some audio software

I see no problem in your code.

Most likely the problem is in your CSV file. I think it was saved with encoding UTF-8 with BOM-signature. Try to change one line in your code:

The question is, when I use the "getkeys" method, it always returns an empty list.

event.getKeys() is an instantaneous check of the keyboard buffer. It will always return an empty list unless a key has been pressed since the last call to the function. So it is usually called in a loop, often once per screen refresh, to detect when a keypress has occurred. The result is that it will return an empty list many times, until a key is actually pressed.

But PsychoPy's event module has actually been deprecated in favour of the newer Keyboard class. This has the advantage that it returns the actual time that the key was pressed: when using event.getKeys(), one just infers the reaction time from the time that the function was called. This is some time after the key is pressed - even if you are calling the function on every screen refresh (typically at 60 Hz), this limits the time resolution to 16.7 ms. The Keyboard class, meanwhile, runs in a separate process and continually checks the keyboard hardware, allowing a precision more at the millisecond level.

Documentation on the Keyboard class is here: https://www.psychopy.org/api/hardware/keyboard.html

You can do this with zip and then flatten it to get items back. You can zip the 2 lists element-wise (corresponding elements of both lists added to a tuple together.

Post that you can use [item for sublist in list for item in sublist] which lets you flatten and break the tuples into corresponding items.

List this -

While it would be possible to open the output file multiple times and add data to it (by using a mode argument of 'a' for "append" instead of "w" for normal (over) writing—see the open() function documentation—in this case it would be less awkward to just open the output file once and leave it that way while appending data to it from the input files, one-by-one.

This is what I mean: (Note: no attempt was made to sort the data because you haven't specified how you wanted that done.)

If you are facing difficulties with PsychoPy, which is really just a set of modules with the full power of Python behind it, then you are certainly going to be frustrated with options like PsyToolKit, which, being browser-based, is never going to be able to give you the level of control you need over your system.

OpenSesame is Python (and often PsychoPy) based under the hood. Paradigm seems to have a nice interface but to add more sophisticated functionality, also uses Python scripting (although seems to still be at version 2.6). I'm not that familiar with SuperLab, but it has been around a long time, and does seem to be focussed on lab equipment use.

But it seems that your issues are more to do with working in parallel with other computer processes rather than PsychoPy per se (which doesn't impose any limitations other than those inherent in Python itself). As noted above, the alternatives you list either won't give the level of control you need, or will also require Python for hardware interfacing.

You might be best to describe specific issues and seek help with those. Many other people have used PsychoPy and Python to display stimuli and record EEG - this task is in no way insurmountable.

I'm experiencing the same issue but found that this is only the case on my computer. When I use the exact same code on a different computer, everything appears normal.

I'm not sure if it's available back in version 1.84 (or why you need to use such an old version of PsychoPy), but the general advice now would be to avoid using the RatingScale class and instead use the simpler but more extensible Slider class.

Regardless, what I'd suggest is that you avoid modifying that object by adding extra markers and so on. Instead create and draw a separate rectangular ShapeStim object: https://www.psychopy.org/api/visual/shapestim.html

Set the vertical position of that rectangle to be a constant, either above, below, or behind the slider as required. On each frame, set its centre to correspond to the marker value of the slider or rating scale. Monitor the mouse scroll wheel, and use that to control the length of the rectangle, to indicate the uncertainty width.