When creating a new conecatch model, one of the required inputs states “Set the maximum number of interaction terms: e.g. 3=Red*Green*Blue.” There are different inputs for “Interaction levels” and “Polynomial level.”
What’s the difference between interaction levels and polynomial level? Do the interaction terms correspond to the number of color cone types in the visual system of interest, the number of color channels in my images, or something else entirely? I’m modeling a UV sensitive avian visual system, so should I set one of these values to 4? Thank you!
There’s usually no need to change the default value, as these settings simply determine the complexity of the model used to map cone catch estimates to camera reflectance. See the 2015 MICA paper for details.