Welcome the latest addition to the tidyverse: duckplyr!
https://duckplyr.tidyverse.org/
The duckplyr package aims to be a fully compatible drop-in replacement for dplyr, using duckdb when it can for performance, and always giving the same results as plain dplyr. If you haven't seen it yet, give it a try and tell us how it works for you!
@kirill I will allow telemetry so you can see some nightmares!
@kirill Nice! Thanks for this contribution and for breakfast the other day
@LuisDVerde My pleasure, thanks for the company!
@kirill instead of "fully compatible drop-in replacement for duckplyr" I presume you mean "fully compatible drop-in replacement for dplyr".
@kirill edit noted!
@kirill what are the most likely cases where you'd expect better performance with duckplyr than dplyr? Data that is large compared to available RAM?
@peter_ellis @kirill From my basic reading, duckplyr makes use of some of the duckdb api functions that wouldn’t translate to dplyr/dbplyr using a DBI overlay. This makes it more performant.
@kirill can't wait to try it out. Found this line saying that factor column is not supported yet. What's the current practice to, for example, ggplot-ing/visualize ordered row? Since I usually use factor/levels here.
@kirill how is duckplyr a drop in replacement for duckplyr?
@kirill where's the hex logo? I have the sticker but the image isn't online :(