decdc and fir_nodelay NAs in R

[dasweeney4423]

I get a return of all NAs with decdc when the matrix contains NAs. However, fir_nodelay (and thus anything that uses it) is what gives errors with NAs:

Error in na.omit.ts(x1, filt/a[1], sides = 1) :
time series contains internal NAs
Called from: na.omit.ts(x1, filt/a[1], sides = 1)

So, two different issues with different things that are happening due to NAs. I just tried these things in matlab for this same tag and the code runs and NAs are only spit out for the regions where the tag data has NAs... looks like this is just an R issue. Checking into the code behind the decdc, the issue with spitting out all NAs is that the convolution function(s) (both pracma and signal versions) can't take NAs and if the input has them, they return only NAs. I am not sure about why fir1 in R is unhappy