Feather, Parquet and the need for space and speed

[NikosAlexandris]

I hope it's OK putting this here, just as a reference or even yet another invitation for an open exchange of information, ideas and experiences. Searching for performace and CSV and the sorts of keywords that relate to Miller, I came across two recently published articles:

• https://towardsdatascience.com/stop-using-csvs-for-storage-this-file-format-is-150-times-faster-158bd322074e (discusses about https://github.com/wesm/feather)
• https://towardsdatascience.com/csv-files-for-storage-no-thanks-theres-a-better-option-72c78a414d1d (discusses about https://github.com/apache/parquet-format)