The Same 540K Rows Can Be 3.5MB or 102.7MB: How to Choose Between CSV, Excel, JSON and Parquet
I saved 540,000 rows of transactions from a UK online store as CSV, JSON and Parquet. The files came out at 45.0MB, 102.7MB and 3.5MB, and with CSV and JSON the types of the date and customer ID columns changed the moment I read them back. Using all 151 Pokémon from PokeAPI, I also cover the pitfalls of flattening nested JSON into a table and monthly partitioning, all with real measurements.