I see .when coming in a future version of dplyr. This'll save so many clunky ifelse() statements! #rstats
20
57
10
538
Replying to @sharoz
Fantastic. Proudly reinventing @Rdatatable and calling it innovation. #Rstats 🤦

Jul 13, 2022 · 7:00 PM UTC

9
3
33
Replying to @eddelbuettel @rdatatable
People can have syntactic preferences. We can let people enjoy features being added to their preferred library.
35
Replying to @eddelbuettel @sharoz @rdatatable
At least they didn't do it with a wrapper, I guess.
3
Replying to @eddelbuettel @sharoz @rdatatable
Hi. Conditional modification of anything is standard in any data thingie you can think of. I love data.table for its elegance, speed and simplicity, but not for its readability or expressive power. I think there are tradeoffs when writing any software. We are looking at one.
6
Replying to @eddelbuettel @sharoz @rdatatable
I think it makes sense for dplyr and data.table to borrow good ideas from each other. They also mention the data.table version in the PR. Here's an example of data.table doing it as well. github.com/Rdatatable/data.t…

Implement DT[, across(.SD, fun1, fun2, fun3), by=group] · Issue #4970 · Rdatatable/data.table

Inspired by dplyr::across and triggered by JuliaData/DataFrames.jl#2725 (comment) Instead of : DT[, unlist(lapply(.SD, function(x) c(max=max(x), min=min(x)))), by=group] it could be DT[, across(.SD...

github.com
26
Replying to @eddelbuettel @sharoz @rdatatable
I think I understand the motivation of @coolbutuseless's tweet now. nitter.vloup.ch/coolbutuseless/s…
14
Replying to @eddelbuettel @sharoz @rdatatable
Bad take; more options and ways of doing things is a good thing.
10
Replying to @eddelbuettel @sharoz @rdatatable
Difference being that you can actually read the dplyr version 🧌
6
Replying to @eddelbuettel @sharoz @rdatatable
This tweet made me stop using datatable
2
Replying to @eddelbuettel @sharoz @rdatatable
Who actually said it's innovation?
1
23