Data architecture - Format
Parquet
A compact column-based file format designed for storing and analysing large datasets
What's it for?
Stores large analytical datasets compactly so only the needed columns can be read efficiently.
For example…
A retailer stores several years of sales in Parquet files so analysts can read only the parts of the data they need instead of loading the whole file every time
A large dataset written as Parquet is much smaller and faster to analyse than an equivalent collection of plain CSV files