Compute profit margin per category with groupby

Question

Finance wants a margin per category: total profit divided by total revenue, not the average of the row-level margins.

Write `margins(df)` that returns a Series indexed by category with the margin as a percentage rounded to 2 decimals, sorted from highest to lowest margin.

Accepted Answer

Step 1: Sum revenue and cost per category *first*. Ratios of sums and averages of ratios are different numbers, and finance wants the former. Step 2: The margin formula `(revenue - cost) / revenue * 100` is applied once per category on the aggregated totals. Step 3: `round(2)` matches reporting precision and `sort_values(ascending=False)` puts the healthiest category first.

Compute profit margin per category with groupby

More Pandas — GroupBy questions

Asked at Deloitte