#64MediumFiles & Modules· Working with Tabular Data

Pandas-Style Sales Analysis

A sales export arrives as CSV lines. Before reaching for a data-analysis library, do the grouping by hand. Write `sales_summary(rows)`. `rows[0]` is a header that contains (in any order) the columns `region`, `product`, `revenue` and `cost`. Every other row is one sale. Fields are separated by commas and never contain commas themselves. - A row is **dropped** if its `revenue` or `cost` is empty or not a whole number (an optional leading `-` is allowed). - For the remaining rows, group by `region` and add up revenue and profit (`revenue - cost`). Return `{"rows_used": <int>, "rows_dropped": <int>, "by_region": [[region, revenue, profit], ...]}` with regions sorted by profit, highest first, and by region name when profits are equal.

Input format

`rows`: a list of CSV lines, the first being the header.

Output format

A dictionary with keys `rows_used`, `rows_dropped` and `by_region`.

Examples

Input: rows = ["region,product,revenue,cost","north,pens,120,70","south,pens,90,40","north,ink,80,100"]

Output: {"rows_used":3,"rows_dropped":0,"by_region":[["south",90,50],["north",200,30]]}

North: revenue 200, profit 50 - 20 = 30. South: revenue 90, profit 50. South ranks first.

Input: rows = ["revenue,cost,region,product","100,,east,tape","abc,10,east,glue","50,20,west,clips"]

Output: {"rows_used":1,"rows_dropped":2,"by_region":[["west",50,30]]}

The columns are in a different order. Two rows have a missing or invalid number and are dropped.

Constraints

1 <= len(rows) <= 10^4
Every data row has as many fields as the header

Study first

Files and Modules in Python

#python
#csv
#data-analysis
#dictionaries
#sorting