Can toolbox process continuous predictor? #137
-
Hi dear Benedikt, Thank you for developing this fantastic toolbox! I'm a postgraduate student, and I'm curious about whether this toolbox can be used to process continuous predictors. Specifically, when an event occurs, I would like to compute a corresponding continuous variable. I'm interested in understanding how the beta value changes in relation to the variable's values. As far as I know, it seems that the toolbox currently allows the categorization of predictors into several conditions. If my understanding is correct, could you please provide any suggestions on how to handle continuous variables ? I hope I have expressed my question accurately. Thanks for your valuable contribution! Best regards, |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
Hi Eve, that is absolutely possible. You can define intercept terms, categorical predictors (factors), continuous linear predictors, or continuous nonlinear (spline) predictors and these can of course also be combined. A continous predictors would be a tone displayed at variable loudness levels. You find a tutorial (for the case of nonlinear continuous predictors) here: https://www.unfoldtoolbox.org/toolbox-tut03.html A different issue are time-continous predictors, that is, predictors that extend through time (e.g. a long audio waveform rather than a discrete event of a single tone). i dont think you mean this but you find the tutorial on that here: https://www.unfoldtoolbox.org/toolbox-tut08.html Cheers, Olaf |
Beta Was this translation helpful? Give feedback.
Hi Eve,
that is absolutely possible. You can define intercept terms, categorical predictors (factors), continuous linear predictors, or continuous nonlinear (spline) predictors and these can of course also be combined. A continous predictors would be a tone displayed at variable loudness levels. You find a tutorial (for the case of nonlinear continuous predictors) here: https://www.unfoldtoolbox.org/toolbox-tut03.html
A different issue are time-continous predictors, that is, predictors that extend through time (e.g. a long audio waveform rather than a discrete event of a single tone). i dont think you mean this but you find the tutorial on that here: https://www.unfoldtoolbox.org/toolbox-…