You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A dictionary of extra information provided to a trial handler does not appear when saving using the experiment handler
Expected Behaviour
As when new loops require creating new data types at the experiment level, I expect loop-level metadata (i.e., experiment phase, a name for the loop) to be incorporated into the data returned in the wide-text output.
Steps to Reproduce
Define an experiment handler (with or without extraInfo)
Define a trial handler with extraInfo
Add the trial handler to the experiment handler
Run experiment and observer output
Additional context
This is an aside, but the workflow around the experiment handler and subordinate trial handlers has been difficult to intuit. I am comfortable reading man pages and python source, but even with the demo code provided on GitHub things like "you must add loops as they are needed, rather than add everything to the experiment handler at the start of the experiment" are not obvious until you bash your head against trial indices not registering properly with the experiment handler for an hour or so. I assume the focus is on getting people to use the builder, but just a bit more documentation around these powerful and useful API tools would go a long way, I think.
The text was updated successfully, but these errors were encountered:
crcox
added
the
🐞 bug
Issue describes a bug (crash or error) or undefined behavior.
label
Nov 25, 2023
A dictionary of extra information provided to a trial handler does not appear when saving using the experiment handler
I've never had the need to do this so we didn't add that functionality. To me the extraInfo that was originally provided in TrialHandler was superseded by the one in ExperimentHandler
"you must add loops as they are needed, rather than add everything to the experiment handler at the start of the experiment" are not obvious until you bash your head against trial indices not registering properly with the experiment handler for an hour or so.
That isn't by design and I didn't know it was the case ;-)
In general, yes, I do strongly recommend you use Builder instead of hand-writing code. There really isn't a good reason to spend hours banging your head against this, and I think hand-written code usually has more problems than the builder-generated code. If only I could design a study to show that! 🤣
PsychoPy Version
2023.1.2
What OS are your PsychoPy running on?
Windows 10
Bug Description
A dictionary of extra information provided to a trial handler does not appear when saving using the experiment handler
Expected Behaviour
As when new loops require creating new data types at the experiment level, I expect loop-level metadata (i.e., experiment phase, a name for the loop) to be incorporated into the data returned in the wide-text output.
Steps to Reproduce
extraInfo
)Additional context
This is an aside, but the workflow around the experiment handler and subordinate trial handlers has been difficult to intuit. I am comfortable reading man pages and python source, but even with the demo code provided on GitHub things like "you must add loops as they are needed, rather than add everything to the experiment handler at the start of the experiment" are not obvious until you bash your head against trial indices not registering properly with the experiment handler for an hour or so. I assume the focus is on getting people to use the builder, but just a bit more documentation around these powerful and useful API tools would go a long way, I think.
The text was updated successfully, but these errors were encountered: