Adding ICA mixing matrix to a different eeg_lst

[stonekate]

Hey Bruno (or whoever's reading), I have some data where I need to run the ICA on data that's filtered differently to the data I want to do analysis on (long story but I can explain if it helps). It's apparently something that is normal to do - my colleague does it, although in ERPlab. In eeguana, I tried doing it something like:

dat1 <- eeg_filt_band_pass(dat, .freq = c(1, 30))

dat1_ica <-eeg_ica(dat1)

dat2 <- eeg_filt_band_pass(dat, .freq = c(0.01, 30))

dat2$.ica <- dat1$.ica

dat2 <- as_eeg_ica_lst(dat2) # error: could not find function "as_eeg_ica_lst"

dat2_icaed <- dat2 %>%  eeg_ica_keep(-13) %>%  as_eeg_lst() # error: no applicable method for 'eeg_ica_keep' applied to an object of class "eeg_lst"

I copied this out of your source code, but this last step doesn't work because as_eeg_ica_lst() doesn't seem to be a function. But then eeg_ica_keep() doesn't work if it's not an eeg_ica_lst. I can't find how as_eeg_ica_lst() is defined in your code. Anyway, any ideas on how I would do this if not this way?

Thanks!

[bnicenboim]

yeah, the function exists but it's not exposed.
you can do:

eeguana:::as_eeg_ica_lst.eeg_lst(dat2)

I'll open an issue to make all this more straightforward with a eeg_add_ica_mat() function.

[stonekate]

It works! And it corrected the blinks nicely. Thanks!