wavenumbers subscript out of bounds error

[davidbenncsiro]

I have a file that yields a wavenumbers subscript out of bounds error:

opusreader::opus_read(path)
# Error in wavenumbers[[i]] : subscript out of bounds
# [1] "Error in wavenumbers[[i]] : subscript out of bounds\n"
# attr(,"class")
# [1] "try-error"
# attr(,"condition")
# <simpleError in wavenumbers[[i]]: subscript out of bounds>

I debugged the code and found that it failed here:

https://github.com/pierreroudier/opusreader/blob/main/R/opus_read_raw.R#L851

The code is:

  spc_m <- setNames(
    lapply(
      seq_along(spc_m),
      function(i) {
        colnames(spc_m[[i]]) <- round(wavenumbers[[i]], 1)
        rownames(spc_m[[i]]) <- unique_id
        spc_m[[i]]
      }
    ),
    names(spc_m)
  )

seq_along(spc_m) yields a sequence from 1 to 7. In this particular case, wavenumbers[[6]] has data, but wavenumbers[[7]] does not exist.

I'm talking with my CSIRO collaborators re: whether I can make the file in question available to you for testing. It's just one file out of almost 1000 and this is the only one that fails. So, whether it's a problem with the code or the file is another question, but it appears to have the right length.

[davidbenncsiro]

I have noticed that with a "healthy" file, seq_along(spc_m) returns 1..6.

Could seq(wavenumbers) be used in place of seq_along(spc_m)?

This may not solve the underlying problem if there is a mismatch between the length of spc_m and wavenumbers however.

In the case of the file that yields an error, while wavenumbers has length 6, spc_m has length 7, but the 7th element is an empty matrix.

[davidbenncsiro]

@pierreroudier I said I would provide the file in question that yields the error. Apologies for the delay.

I have attached it here, zipped:

265_Yanco-Puck_Wt-Invenio_a_A3.zip

[davidbenncsiro]

opusreader2 also appears to fix this bug @philipp-baumann.