Problem
Kaggle provided me with data that it scraped from Wikipedia. The resulting spreadsheet was a disaster. Rows didn't line up. Information was obscured by improperly sized columns. Some cells were blank. It wasn't reasonably usable in the state it was in.
Process
- I rearranged the columns so the data reads left to right in a way that makes sense.
- I fixed the alignment and adjusted the column width so the cut-off text was visible.
- Blank cells became 'N/A' so I could tell the difference between missing data and empty space.
- I used REGEXREPLACE, TRIM, and SUBSTITUTE formulas to strip clutter while keeping the actual source citations intact.
- Documented each change in a Data Cleaning Log.
Result
The dataset is now not only easy to read, but ready to use. Source markers were preserved for traceability.
Want to explore the spreadsheet for yourself?
Find it here!Screenshots
This is the original dataset provided by Kaggle, as scraped from Wikipedia.
Here is the final cleaned dataset with nice clean columns and formatting!
This is my Data Cleaning Log with metrics, the issues I found, and improvements I made.
Here's an example of how I use formulas!