Python CSV Cleanup & Validation — Demo

The problem

A spreadsheet export can look usable while still containing duplicate orders, missing IDs and invalid amounts. Those errors need to be caught before the data is used in a report or imported elsewhere.

What I built

A small Python workflow that separates accepted orders from records needing review. This demonstration uses fictional data.

How it works

The script trims extra spaces, checks order IDs and amounts, and formats accepted amounts to two decimal places. It writes two CSV files: clean records and rejected rows. Each rejected row keeps its source row number and a reason, so it can be traced back without changing the original input.

What I verified

The 10-row sample produced 6 accepted records and 4 flagged records. Checks covered duplicate and missing IDs, invalid amounts, negative values and non-finite values such as NaN.

The result is a repeatable cleanup step with a clear review list. For a client workflow, I would agree on the data rules and test them against representative files before using it on live data.