Not sure about the API on this because from what I understand from @bkietz, every aggregation function takes a single field ref, but group_by then count doesn't require a field because you're counting how many rows are in each group. (Perhaps more importantly, dplyr::n() takes no arguments, so we don't have a field ref in our expression to start.)

Reporter: Neal Richardson / @nealrichardson
Assignee: Neal Richardson / @nealrichardson

Related issues:

• [R] Use Arrow engine for summarize() by default (blocks)
• [R] Support dplyr::count() and tally() (blocks)
• [C++] Support scalar columns in hash aggregations (was: hash_sum on scalar column segfaults) (is blocked by)
• [R] Initial bindings for ExecPlan/ExecNode (relates to)
• [R] Binding for n_distinct() (relates to)
• [R] Rename na.min_count to min_count and na.rm to skip_nulls (is related to)

PRs and other links:

• GitHub Pull Request #10989

_{Note: This issue was originally created as ARROW-13501. Please see the migration documentation for further details.}