Repository navigation
Need advice on cleaning a dataset #30
Description
So I am working on my first data analysis project and am uncertain how much of the data I should clean. I have the raw data properly organized but alot of rows have blanks. None of the blanks affect any of the fields the data project is concerned so I'm unsure whether or not to keep those entries. I currently have a cleaned dataset and a raw dataset. The cleaned dataset has no rows with blanks or invalid entries, the uncleaned dataset keeps those rows in (minus the invalid entries). It's alot of entries thrown out if I use the clean dataset so I'm unsure what to do. Would greatly appreciate any advice.
Activity
adamfige commentedon Aug 29, 2025
"Good point! Since the blanks don’t affect your analysis fields, I think it’s fine to keep those rows. Keeping both raw and cleaned versions is a smart approach."
Nikith20072002 commentedon Oct 6, 2026
Analyze the missing data. If more than one value is missing in each row maybe you can consider dropping the rows. If that helps to clean most data leaving only around 2%(example value can be more or less but not more than 5%) the filling those with mode or median will be helpful. But if it is a small dataset drop all null rows.