coderfender commented on code in PR #2817:
URL: https://github.com/apache/datafusion-comet/pull/2817#discussion_r2560277019
##########
native/core/src/execution/planner.rs:
##########
@@ -1893,19 +1893,24 @@ impl PhysicalPlanner {
let child = self.create_expr(expr.child.as_ref().unwrap(),
Arc::clone(&schema))?;
let datatype =
to_arrow_datatype(expr.datatype.as_ref().unwrap());
let input_datatype =
to_arrow_datatype(expr.sum_datatype.as_ref().unwrap());
+ let eval_mode = from_protobuf_eval_mode(expr.eval_mode)?;
+
let builder = match datatype {
DataType::Decimal128(_, _) => {
let func =
AggregateUDF::new_from_impl(AvgDecimal::new(datatype, input_datatype));
Review Comment:
@martin-g , EvalMode is pretty much redundant for integer data types because
the running sum is stored as a float and Spark considers infinity as a valid
value.I will be raising a new PR to support ANSI mode for decimal inputs since
it the changes are in a different file all together
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
