Advice on Reducing Results File Size

[cdfelton]

Dear Jeremy,

Thank you so much for all your work on PennController!

I am currently encountering a similar issue to many others after the farm migration where my old code simply produces results files that are too large. I only need to actually record the data from lines where the results parameter is “Final”, which would greatly reduce the file size. Since I inherited that part of this code from my lab, I do not know how to only record those lines while still making sure that all the results are saved to a line that also includes the item information from my stimuli csv that I’ll need later to analyze it. Is there a change I can make to my code that will only record the “Final” parameter items but keep their formatting the same?

Demo Link:
https://farm.pcibex.net/r/BrnGEO/

Best,
Casey

[Jeremy]

Dear Casey,

Some things you can do:

– if you don’t need the starting and ending time of your Header trials, add .noTrialLog() after the closing parenthesis of your Header(...) (before or after .log(...), it shouldn’t matter)

– more generally, if you don’t need the starting and ending time of some trial, add .noTrialLog() after the closing parenthesis of its corresponding newTrial(...) command

– if you are not interested in the participants’ answers to the practice trials, consider remove the .log(...) commands on the elements included in those trials (Key elements, TextInput elements, etc.)

– for your Textnput elements, you may want to consider using .log("final") so as to only get the “Final” line, and not the “first” and “validate” ones

– you may or may not want to delete some of the .log commands on your Key elements, depending on how crucial it is for response time calculations to have lines reporting those keypresses in your results file

Most importantly, make sure to take one or several testruns after editing your project to verify the integrity of the results files it produces, and apply the analyses you would normally apply so you’re sure you’re not missing any crucial bit of information

Jeremy

[cdfelton]

Dear Jeremy,

I so appreciate your prompt reply, your suggestions worked perfectly and seem not to have impacted the integrity of my results file at all besides, if anything, simplifying data analysis! I will be taking your advice and doing several full length test runs before re-publishing, but it seems like each participant will only be ~400 lines as opposed to ~2400.

Best Wishes and Thank You,
Casey

[Author]

Posts