alamb commented on code in PR #5554:
URL: https://github.com/apache/arrow-datafusion/pull/5554#discussion_r1135516150
##########
datafusion/physical-expr/src/aggregate/count_distinct.rs:
##########
@@ -85,64 +86,124 @@ impl AggregateExpr for DistinctCount {
}
fn create_accumulator(&self) -> Result<Box<dyn Accumulator>> {
- Ok(Box::new(DistinctCountAccumulator {
- values: HashSet::default(),
- state_data_type: self.state_data_type.clone(),
- }))
+ use arrow::datatypes;
+ use datatypes::DataType::*;
+
+ Ok(match &self.state_data_type {
+ Dictionary(key, val) if key.is_dictionary_key_type() => {
+ let val_type = *val.clone();
+ match **key {
+ Int8 => Box::new(
+
CountDistinctDictAccumulator::<datatypes::Int8Type>::new(
+ val_type,
+ ),
+ ),
+ Int16 => Box::new(
+
CountDistinctDictAccumulator::<datatypes::Int16Type>::new(
+ val_type,
+ ),
+ ),
+ Int32 => Box::new(
+
CountDistinctDictAccumulator::<datatypes::Int32Type>::new(
+ val_type,
+ ),
+ ),
+ Int64 => Box::new(
+
CountDistinctDictAccumulator::<datatypes::Int64Type>::new(
+ val_type,
+ ),
+ ),
+ UInt8 => Box::new(
+
CountDistinctDictAccumulator::<datatypes::UInt8Type>::new(
+ val_type,
+ ),
+ ),
+ UInt16 => Box::new(CountDistinctDictAccumulator::<
+ datatypes::UInt16Type,
+ >::new(val_type)),
+ UInt32 => Box::new(CountDistinctDictAccumulator::<
+ datatypes::UInt32Type,
+ >::new(val_type)),
+ UInt64 => Box::new(CountDistinctDictAccumulator::<
+ datatypes::UInt64Type,
+ >::new(val_type)),
+ _ => {
+ // just checked that datatype is a valid dict key type
+ unreachable!()
Review Comment:
It would be preferable to fix this prior to merge, but I think it would also
be fine to merge as is
--
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]
