group_by.dtplyr_step {dtplyr} | R Documentation |
These are methods for dplyr's group_by()
and ungroup()
generics.
Grouping is translated to the either keyby
and by
argument of
[.data.table
depending on the value of the arrange
argument.
## S3 method for class 'dtplyr_step' group_by(.data, ..., .add = FALSE, add = deprecated(), arrange = TRUE) ## S3 method for class 'dtplyr_step' ungroup(.data, ...)
.data |
|
... |
In |
.add, add |
When This argument was previously called |
arrange |
If |
library(dplyr, warn.conflicts = FALSE) dt <- lazy_dt(mtcars) # group_by() is usually translated to `keyby` so that the groups # are ordered in the output dt %>% group_by(cyl) %>% summarise(mpg = mean(mpg)) # use `arrange = FALSE` to instead use `by` so the original order # or groups is preserved dt %>% group_by(cyl, arrange = FALSE) %>% summarise(mpg = mean(mpg))