What counts as erroneous data, as this is not properly explained. So far I’ve understood that something like -1 in the month column would be erroneous, and to then drop that row as a result, but the uncleaned TSV does not have any occurrences of this (and hence makes it harder to test for). On top of that, are these errors only limited to fields that are only meant to contain numbers? I don’t find it particularly feasible to test the fields that arent limited to numbers (some of which have paragraphs in them) for this sort of data. In a previous HelpOSTS question someone also asked about posting errors to stdout/stderr. In this case, will the files that our preprocess script is tested against contain things that would be expected to cause bash errors? Or is the erroneous data simply something like a month that doesnt exist (e.g. -1 as a month)? I ask this question purely because in my codes current state the output to stdout is equivalent to the cleaned file that was provided, even without any code to deal with errors. A tip for a future assignment would also be if the uncleaned data we were given to test it with actually had the erroneous data we were meant to fix as well so we could test that out, instead of asking for something but not giving us a data file that we can appropriately test with.