I'm wondering how people feel about returning more information with the time_generalization function. Right now, it runs cross validation, fits a bunch of models, generalizes all those models, and then throws away all of that information except for a single averaged score matrix.
It'd be useful to have access to the coefficients in order to see which regions of the brain are stable across time (I'm doing this with ECoG). It'd also be useful to have the model coefficients and scores across all of the CVs, in order to get an idea for how stable these values are.
Right now I've just been modifying this function locally to return these things, but I wonder if they wouldn't be useful to others as well...
I'm wondering how people feel about returning more information with the time_generalization function. Right now, it runs cross validation, fits a bunch of models, generalizes all those models, and then throws away all of that information except for a single averaged score matrix.
It'd be useful to have access to the coefficients in order to see which regions of the brain are stable across time (I'm doing this with ECoG). It'd also be useful to have the model coefficients and scores across all of the CVs, in order to get an idea for how stable these values are.
Right now I've just been modifying this function locally to return these things, but I wonder if they wouldn't be useful to others as well...