How to set to NAs some segments of a channel

[bnicenboim]

See also #167

[bnicenboim]

Example:

eeg_fake <- eeg_lst(signal_tbl = tibble(.id = rep(1:10, each =100),
.sample = sample_int(rep(1:100, 10),500),
P1 = channel_dbl(rnorm(1000)),
P3 = channel_dbl(rnorm(1000))))

eeg_mutate(eeg_fake, P3 = case_when(.id %in% 1:5 ~ NA,
TRUE ~ P3))

[stonekate]

Thanks for this. The problem I'm having now is that when there are NAs in the signal table, various functions don't work. I believe it's because there's a check somewhere in datatable that is.double == TRUE, and the NAs violate this. So far I've noticed it affects eeg_artif_minmax() and bind(). And probably related functions like eeg_artif_step(). Here's the error message I get:

As a reprex using your code above, I get a shorter version of the error (same problem though I think):

eeg_fake <- eeg_lst(signal_tbl = tibble(.id = rep(1:10, each =100),
.sample = sample_int(rep(1:100, 10),500),
P1 = channel_dbl(rnorm(1000)),
P3 = channel_dbl(rnorm(1000))))

eeg_fake <- eeg_mutate(eeg_fake, P3 = case_when(.id %in% 1:5 ~ NA,
TRUE ~ P3))

eeg_artif_minmax(eeg_fake, .threshold = 150, .window = 150, .unit = "ms", na.rm = TRUE)

[bnicenboim]

the problem with the minmax function is the na.rm argument that doesn't exist. This eeg_artif_minmax(eeg_fake, .threshold = 150, .window = 150, .unit = "ms") works.
And I can also bind two eeg objects with NAs.
bind(eeg_fake,eeg_fake) works as well.

[stonekate]

Oh sorry I left the na.rm in by accident (was just testing). For my non-fake dataset I get the error message, but maybe there's some other problem with my data. I will look into it some more.