That is precisely my issue. The clean data that is provided has 1056 lines, and likewise, the dirty data provided has 1056 lines. In other words, no lines get dropped whatsoever. I HAVE compared the output of my script with the provided clean data file using the vim -d command, and it shows a perfect match between my output and the clean file, despite my output not dropping any rows with erroneous data. The dirty data we are given has no lines that have an invalid month or year for example (and as a result it makes it more difficult to create a script that detects these when the input data would not have any), nor are there any empty cells, etc. The provided dirty data and clean data is insufficient in this sense as we are now forced to test for errors that dont exist in the provided data.