A file contains the value 12,5. Is the comma part of one decimal number or the separator between two fields? The answer belongs to the file’s format, not to the appearance of the number alone.
The pandas read_csv documentation provides separate settings for the field separator and the decimal character. Those choices allow a semicolon-delimited file to use commas inside numeric values without treating every comma as a column boundary.
Consider an invented row: sampleA;12,5;7. If the documented separator is a semicolon and the decimal mark is a comma, the row has three fields. Splitting it at every comma would produce a different interpretation before any calculation began.
Inspect a small known example
Open a copy as plain text and compare the header with several rows. Check quoting too: a separator may appear inside a quoted text field. Then import with explicit settings and inspect both the resulting values and their data types.
Choose one value whose intended interpretation you can verify from the data dictionary. Test that value before processing the complete file. A successful import message means the software accepted an arrangement; it does not prove that arrangement matches the publisher’s meaning.
Record the chosen delimiter, decimal mark and encoding beside the import command. If you later receive a file from another source, revisit those settings instead of assuming that a familiar filename extension guarantees an identical format.
Editorial illustration from the site library.