Closed hadley closed 5 years ago
To be more concrete, you'll now need to explicitly start dplyr pipelines with lazy_dt()
and explicitly force computation with as.data.table()
/as.data.frame()
/as_tibble()
.
I think it should be a fairly simple change to your code, but please let me know if there's anything I can do to help.
Thanks for the note @hadley! Actually I wasn't able to find the function that actually imports dtplyr, so I suspect it might have been in the Imports
field just because of historical reasons. @pbiecek @OlaPaluszynska If either of you could confirm, I'm going to remove dtplyr from Imports
. I've tested it locally and nothing seems affected.
@Yue-Jiang yes it's likely that's a dependency from some early version of randomForestExplainer. Would you check if other packages from the DESCRIPTION are still needed there?
This is a heads up to let you know that I'm planning to release dtplyr 1.0.0 in the near future. This is a complete rewrite of dtplyr so it generates vastly more performant data.table code, but it did require a complete rework of the API so your existing dtplyr code is unlikely to continue to work.