Post #4236021
2025-03-11 17:36 UTC
@ginolhac@fosstodon.org as far as I can tell they're all faster, all have lower memory requirements, but vary in the degree to which they are a true drop-in replacement for your current tidyverse code. I wish there was a way to know in advance which one is most likely to work for me, but on the otherhand, I guess it's not too hard to try out each one 🤷♂️
Replies (2)
-
@ginolhac@fosstodon.org 2025-03-11 18:56
@LeafyEricScott@fosstodon.org @etiennebacher@mastodon.social to me tidypolars covers more functions useful for data analysis. And polars takes 10min to compile but duckdb 40 I discovered 🫨
-
@jimgar@hachyderm.io 2025-03-12 00:11
@LeafyEricScott@fosstodon.org @ginolhac@fosstodon.org They all seem to break in their own hilarious and depressing ways don’t they. I prefer to use the ‘official’ bindings or, in duckdb’s case, just write the SQL myself and pass it to duckdb as a string. If you’re just doing filters, summarising, joining etc, in many cases the syntax is going to be so straightforward that it seems pointless to use dplyr with a backend. Like if you know dplyr already then basic SQL is very easy to grasp. So is Polars.