Normalization organizes information so each fact appears once, either by standardizing inconsistent values or separating related data into structures that can be joined when needed. section 5.13 section 6.8
Standardizing vehicle colors
A parking-violations dataset may record WHITE as WHITE, WT, WT., and WHT. Count distinct values, inspect the most common colors, build a dictionary mapping alternate inputs to canonical names, replace them, and inspect the remaining inconsistencies and errors. section 5.13
Separating products and sales
Keep product details in products_df and transactions in sales_df, set product_id as the index in both, and join the frames. The combined data supports totals by product, income by product, daily income, and product income by day. section 6.8
products_df.join(sales_df)
A concise pandas join makes normalized product and sales data available for subsequent grouping and sorting. section 6.8
