pitrou commented on code in PR #13327:
URL: https://github.com/apache/arrow/pull/13327#discussion_r893249117
##########
python/pyarrow/_compute.pyx:
##########
@@ -2047,6 +2047,60 @@ class RandomOptions(_RandomOptions):
self._set_options(length, initializer)
+cdef class _RankOptions(FunctionOptions):
+
+ _tiebreaker_map = {
+ "Min": CRankOptionsTiebreaker_MIN,
+ "Max": CRankOptionsTiebreaker_MAX,
+ "First": CRankOptionsTiebreaker_FIRST,
+ "Dense": CRankOptionsTiebreaker_DENSE,
+ }
+
+ def _set_options(self, sort_keys, null_placement, tiebreaker):
+ cdef vector[CSortKey] c_sort_keys
+ for name, order in sort_keys:
+ c_sort_keys.push_back(
+ CSortKey(tobytes(name), unwrap_sort_order(order))
+ )
+ try:
+ self.wrapped.reset(
+ new CRankOptions(c_sort_keys,
unwrap_null_placement(null_placement), self._tiebreaker_map[tiebreaker])
+ )
+ except KeyError:
+ _raise_invalid_function_option(tiebreaker, "tiebreaker")
+
+
+class RankOptions(_RankOptions):
+ """
+ Options for the `rank` function.
+
+ Parameters
+ ----------
+ sort_keys : sequence of (name, order) tuples
Review Comment:
> If it makes it easier I can add this back into the C++ implementation.
I would rather not. The array-specific kernel for sorting is not a very
user-friendly API pattern. Let's just make `RankOptions` usable enough in
Python :-)
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
