Writing Parquet files using Haskell

(datahaskell.org)

52 points | by cosmic_quanta 2 days ago

3 comments

  • mugul 49 minutes ago
    Looks like a pretty cool community building great tools with care. Using a functional language for data transforms sounds like a sane idea, haven't played around with it yet but it's definitely on my list now.

    However they claim using Haskell for data science is "fast", which doesn't really mean anything until you have numbers to show. A little benchmark with pandas and polars wouldn't hurt I guess.

    • epgui 26 minutes ago
      Haskell tends to be C-fast.
      • veltas 12 minutes ago
        That's not what I see reported, they say Haskell tends to have bad memory layout generally and takes a 5x or so hit to performance.
  • Quothling 3 hours ago
    I'd personally rather see Haskell become part of the options for https://arrow.apache.org/, but this is still a cool project.
  • pandinus 5 hours ago
    [flagged]
    • kccqzy 4 hours ago
      Funny how you mention it is a functional programming torture course, but the Parquet writer introduced in this article is written in a highly imperative style. I have not seen this much IORefs and MutableByteArrays in recent memory.
      • okasaki 1 hour ago
        That's just what you have to do to get any kind of performance in Haskell because functional programming is pointless.
    • 4b11b4 5 hours ago
      What! I loved my func prog class with Haskell

      Should be mandatory