A sales data analysis
The problem
A shop exports a year of messy sales records. Clean them, answer which products, regions and months drive sales, and write a short report with the cleaning log.
Requirements
- A cleaning step that fixes spaces, case and dates, drops broken rows and duplicates, and logs every count
- Monthly totals with growth and a 3-month moving average
- Totals and averages by region and product, joined with a product table for categories
- A report that states the findings, the data dropped and the limits
Technologies
- Python 3
- csv and datetime
- Dictionaries and sorting
- Descriptive statistics
Step by step
- Write the questions the shop wants answered and profile the raw file.
- Build and test the cleaning step, logging each change.
- Group, join and compute the summaries and the monthly series.
- Write the report with tables, findings and limits.
Expected result
A pipeline that rebuilds the whole report from the raw file in one run.
How it is evaluated
- Every cleaning decision is logged with counts
- Totals match hand calculations on a sample
- Findings answer the stated questions
- The report states its limits
Ways to improve it
- Add a text chart of monthly sales
- Flag anomalous days
- Compare each month with the same month last year