News for hackers. Stuff that matters.

A comment on Using machine learning to choose compression algorithms.

Re:Using machine learning to choose compression algorithms by mattnewport on Sunday December 08, 2019 @05:21PM (#21738810)
From the data shown it's not obvious that some of the options would ever be best, would be interesting to see if any of the non csv options ever beat the csv options and for what type of data.
[ Parent | on HN ]
Re:Using machine learning to choose compression algorithms by kootenpv on Sunday December 08, 2019 @05:29PM (#21738867)
I can give you this... the larger the data the more useful parquet and compression on it will be...
[ Parent | on HN ]
Re:Using machine learning to choose compression algorithms by mattnewport on Sunday December 08, 2019 @06:19PM (#21739159)
From the brief description it seems like Parquet is just a weaker version of what a decent LZ style compressor will do anyway
[ Parent | on HN ]