mostPopularProduct.py

import json
import pandas as pd

with open('input_p2.json') as json_file:
    df = pd.DataFrame(json.loads(line) for line in json_file)
    groupByProductId = df.groupby('product_id')
    countOfDistinctUserId = groupByProductId['user_id'].nunique().sort_values()
    print("Most popular product(s) based on the number of purchasers: ",
          countOfDistinctUserId.loc[countOfDistinctUserId == countOfDistinctUserId.max()].index.tolist())
    sumOfQuantity = groupByProductId.sum()
    print("Most popular product(s) based on the quantity of goods sold: ",
          sumOfQuantity.loc[sumOfQuantity['quantity'] == sumOfQuantity['quantity'].max()].index.tolist())