{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreiaxwu7atvyto2kycwcjlyzlm36nbrysuuwiho472iiatw42mkxgom",
    "uri": "at://did:plc:pi6woz4d47bkuws673w2il2r/app.bsky.feed.post/3mnmpcieldbu2"
  },
  "path": "/t/rfc-sibyl-time-series-analysis-in-haskell/13823?page=2#post_23",
  "publishedAt": "2026-06-06T12:27:50.000Z",
  "site": "https://discourse.haskell.org",
  "textContent": "Bumping this thread for ZuriHac 2026.\n\nI am planning to start up work on Sibyl again (I got super bogged down with university work) after ZuriHac ends and are currently wondering if it would be possible to get an idea of how people feel about time series work in Haskell. If you’re a data scientist or statistician – please, please comment as to how a library like this would save you time.\n\nSibyl, and a lot of dataHaskell right now, is a question of “How can we best impact users coming from outside of Haskell and within Haskell without just duplicating `pandas`, `statsmodels` or existing tooling in other ecosystems?” I think the correct answer is always to keep things in-line with Haskell, idiomatic and strongly typed with few runtime errors, while also trying to make it approachable for non-Haskellers or seasoned category theory vets. The UX is extremely important. And, I think, Haskell can be quite a bit more performant than a lot of these libraries are in Python. So the correct balance is between convenience/UX and performance/complexity.\n\nAnd if you’re at ZuriHac, come see us in Room 4.113 in the left corner and tell us what you think.",
  "title": "[RFC] Sibyl: Time Series Analysis in Haskell"
}