Hi all,

I am not sure, whether this example should work or works as intended

df <- DataFrame(one = c(1,2,3), two = c(NA,NA,NA))
df[is.na(df)]
Error: subscript is a logical vector with out-of-bounds TRUE values
In addition: Warning message:
In NSBS(i, x, exact = exact, strict.upper.bound = !allow.append,  :
  subscript is an array, passing it thru as.vector() first

The logical matrix seams to be reasonable valid

is.na(df)
      [,1] [,2]
[1,] FALSE TRUE
[2,] FALSE TRUE
[3,] FALSE TRUE

In comparison using data.frame

df2 <- as.data.frame(df)
df[is.na(df)]
[1] NA NA NA

Any advice or suggestions? Did I miss something obvious? Is the use case I described discouraged somehow?

Best

Hi Felix,

My guess is that the [ method for DataFrame objects doesn't know how to handle a logical matrix subscript in a way that mimics what happens with ordinary data frames or matrices. Note that in such case df[m] can simply be obtained with as.matrix(df)[m] so the workaround is pretty straightforward. The point could be made that the error message produced by df[m] doesn't really help though.

Might be worth opening an issue here https://github.com/Bioconductor/S4Vectors/issues

Thanks, H.