popularProducts.py

import json

# This is probably super straight forward with either pandas
# or a simple sqlite db. Showing how I would implement it without those
# in this sample.


highest_unique_sales = 0
highest_unique_sales_pids = []

highest_sales_qty = 0
highest_sales_qty_pids= []

by_counts = {}
by_users = {}

fh = open('SWE sample data - Q2 data.csv')
for line in fh:
  data = json.loads(line)
  uid = str(data["user_id"])
  pid = str(data["product_id"])
  qty = int(data["quantity"])

  if pid in by_users:
    if uid not in by_users[pid]:
      by_users[pid][uid] = 1
    else:
      by_users[pid][uid] += 1
  else:
    by_users[pid] = {
      uid: 1
    }

  if pid in by_counts:
    by_counts[pid] += qty
  else:
    by_counts[pid] = qty

  # By doing the aggregation during the file read process,
  # complexity reduces since we arent doing multiple loops anymore.
  unique_sales = len(by_users[pid])
  if unique_sales > highest_unique_sales:
    highest_unique_sales = unique_sales
    highest_unique_sales_pids = [pid]
  elif unique_sales == highest_unique_sales and pid not in highest_unique_sales_pids:
    highest_unique_sales_pids.append(pid)


  if by_counts[pid] > highest_sales_qty:
    highest_sales_qty = by_counts[pid]
    highest_sales_qty_pids = [pid]
  elif by_counts[pid] == highest_sales_qty:
    highest_sales_qty_pids.append(pid)

fh.close()

print("Most popular product(s) based on the number of purchasers: [ "+ ",".join(highest_unique_sales_pids)+ " ]")
print("Most popular product(s) based on the quantity of goods sold: [ " + ",".join(highest_sales_qty_pids) + " ]")