Clarify Usage of .BY in Special Symbols Documentation
#6193
Conversation
man/special-symbols.Rd
Outdated
| @@ -66,5 +68,25 @@ DT[{cat(sprintf('in i, .N is \%d\n', .N)); a < .N/2}, | |||
| # .I can be different in j and by, enabling rowwise operations in by | |||
| DT[, .(.I, min(.SD[,-1]))] | |||
| DT[, .(min(.SD[,-1])), by=.I] | |||
|
|
|||
| #These examples illustrate the importance of using copy(.BY) when capturing .BY values in a list outside the data.table context to avoid unexpected behavior. | |||
| dt <- data.table(l = sample(letters[1:5], 100, replace = TRUE), b = rnorm(100)) | |||
There was a problem hiding this comment.
hi, this is a good start, but how about a simpler example, with fewer than 100 rows?
how about 4 rows, data.table(letter=c('a','a','b','b'),number=1:4)
|
Hi @Nj221102, FYI, there's no need to do the "Update" step for each individual PR every time |
Sure will keep that in mind from now on , sorry for the inconvenience caused |
|
@tdhock plz check this out as well. |
| @@ -66,5 +68,25 @@ DT[{cat(sprintf('in i, .N is \%d\n', .N)); a < .N/2}, | |||
| # .I can be different in j and by, enabling rowwise operations in by | |||
| DT[, .(.I, min(.SD[,-1]))] | |||
| DT[, .(min(.SD[,-1])), by=.I] | |||
|
|
|||
| # These examples illustrate the importance of using copy(.BY) when capturing .BY values in a list outside the data.table context to avoid unexpected behavior. | |||
| dt <- data.table(l = letters[1:5], b = rnorm(5)) | |||
There was a problem hiding this comment.
again please use simpler data (1:4 instead of rnorm) as I mentioned in my previous comment
| byg <<- append(byg, .BY) # Incorrect: Missing copy(.BY) | ||
| mean(b) | ||
| }, by = l] | ||
| str(byg) # All elements of 'byg' will contain the .BY value of the last group | ||
|
|
||
| ## Correct usage | ||
| byg <- list() | ||
| dt[, m := { |
There was a problem hiding this comment.
these examples are perhaps too long.
If I understand correctly, the important different is .BY versus copy(.BY) right?
So can you please delete as much other code (:= append, mean) as possible to make the example simpler? ideally one line instead of 6 for both incorrect/correct usage
closes #5389
Description:
This PR updates the
special-symbolsdocumentation to make it clear how to use.BYcorrectly indata.table. The main goal is to help users understand how to avoid issues when using.BYoutside of its immediate context, as discussed in GitHub Issue #5389.Changes:
detailssection to include advice on usingcopy(.BY)when extracting.BYoutside its immediate context.copy(.BY).