Lab: Wrangle some data

Read a CSV of product inventory, filter low-stock items, and write the results to JSON — the complete parse-transform-serialise loop.

You have a CSV export of product inventory. Your job: find items that are running low on stock, add a computed urgency field, and write the results to JSON so a downstream system can process them.

This lab exercises the complete data-wrangling loop: parse → filter → transform → serialise.

The data

sku,name,category,stock,reorder_threshold
A001,Wireless Keyboard,peripherals,14,20
A002,USB-C Hub,peripherals,5,15
A003,Monitor Stand,accessories,32,10
A004,Webcam HD,peripherals,2,10
A005,Desk Lamp,accessories,0,5
A006,Laptop Sleeve,accessories,18,25
A007,Mechanical Keyboard,peripherals,7,20
A008,Mouse Pad XL,accessories,11,10

A product needs reordering when stock < reorder_threshold. The urgency is reorder_threshold - stock — how many units short you are.

Checkpoint 1 — parse the CSV

Read every row into a list of dicts and print the first two:

Python — editable, runs in your browser

Checkpoint 2 — filter and convert types

Find products that need reordering. Remember: every value from CSV is a string, so convert stock and reorder_threshold to integers before comparing:

Python — editable, runs in your browser

Checkpoint 3 — transform and add a computed field

Build a clean output record for each low-stock item. Add the computed urgency field and drop the raw CSV strings in favour of typed values:

Python — editable, runs in your browser

Checkpoint 4 — serialise to JSON

Write the final output as formatted JSON:

Python — editable, runs in your browser

import csv
import io
import json

DATA = """sku,name,category,stock,reorder_threshold
A001,Wireless Keyboard,peripherals,14,20
A002,USB-C Hub,peripherals,5,15
A003,Monitor Stand,accessories,32,10
A004,Webcam HD,peripherals,2,10
A005,Desk Lamp,accessories,0,5
A006,Laptop Sleeve,accessories,18,25
A007,Mechanical Keyboard,peripherals,7,20
A008,Mouse Pad XL,accessories,11,10"""

def parse_inventory(csv_string):
  reader = csv.DictReader(io.StringIO(csv_string))
  return list(reader)

def find_reorder_items(products):
  results = []
  for p in products:
      stock = int(p["stock"])
      threshold = int(p["reorder_threshold"])
      if stock < threshold:
          results.append({
              "sku": p["sku"],
              "name": p["name"],
              "category": p["category"],
              "stock": stock,
              "reorder_threshold": threshold,
              "urgency": threshold - stock,
          })
  results.sort(key=lambda x: x["urgency"], reverse=True)
  return results

def main():
  products = parse_inventory(DATA)
  reorder_items = find_reorder_items(products)

output = {
      "total_products": len(products),
      "items_to_reorder": len(reorder_items),
      "reorder_list": reorder_items,
  }

print(json.dumps(output, indent=2))

main()

The structure mirrors the automation lab: separate functions for parsing, processing, and output. If the CSV later comes from a real file, only parse_inventory() changes — the filtering and serialisation logic stays intact.

Extend it yourself

Group the reorder list by category in the JSON output instead of a flat list.
Add a "generated_at" timestamp field to the output using datetime.datetime.now().isoformat().
Change parse_inventory() to accept a file path and read from a real CSV on disk.

Finished reading? Mark it complete to track your progress.

The data

Checkpoint 1 — parse the CSV

Checkpoint 2 — filter and convert types

Checkpoint 3 — transform and add a computed field

Checkpoint 4 — serialise to JSON

Extend it yourself

On this page