I'm thinking of posting regular code snippets that can be used to build basic psychophysics experiments. These would include elements of simple code design and structure (using for loops, if statements, stimulus presentation, and data recording, etc).
I will start doing this beginning next year (in January) if there is sufficient interest before then. So tell me, what do you think?
No experience wold be necessary. So long as you have access to Matlab, you can download the free PsychToolbox. You could also recommend your own code, or post a solution to a community problem.
I put something very basic together for the MoNoC group that kinda speaks to this: https://github.com/julian-matthews/MoNoC-practice-experiment.
Not sure if many of them picked it up but I found it handy to have around when I was guiding students that were new to programming. Could simply direct them to particular lines.
I'd like to see something similar for analysis scripts perhaps? I know my early analysis scripts were a complicated mess and my current ones could definitely be improved. I really like the idea of R Markdown for that kinda thing but I'm still learning.
Thanks for this Julian, this is great! This is similar to what I had in mind. I'll happily refer people to your project as another example. Maybe we could write a joint project at some point too?
My core idea is slightly different, but it would nicely extend what you've already done above. I was thinking of focusing on making the program highly modular (a lot like the way web design works). This way I can show how to make lots of very small programs that perform a singular function (e.g., add in a fixation cross using a pre-written file, or make the design MOCS, or staircase, etc, using another file). And 'build' the experiment from modules into a single, dense program with only a handful of lines of code that refers to each module along the way.
This program design adds to your suggestion too: We can write analysis scripts in a modular fashion and add them to the end of the program to automate analysis, or write modules for arranging results so they can be exported to dedicated analysis software.
I'll continuously add code into this project over the next few months: https://www.researchgate.net/project/Basic-code-design-in-Matlab-PsychToolbox
and we should chat more if we have ideas that would make the whole thing better! (like making it a project on GitHub for example).
- Badges
- Science topic