[GitHub] [spark] harupy commented on a change in pull request #32245: [SPARK-35142][ML] Fix incorrect return type for `rawPredictionUDF` in `OneVsRestModel`

[GitBox <gi...@apache.org>]

harupy commented on a change in pull request #32245:
URL: https://github.com/apache/spark/pull/32245#discussion_r616310831



##########
File path: python/pyspark/ml/classification.py
##########
@@ -3151,7 +3151,7 @@ def func(predictions):
                     predArray.append(x)
                 return Vectors.dense(predArray)
 
-            rawPredictionUDF = udf(func)
+            rawPredictionUDF = udf(func, VectorUDT())
             aggregatedDataset = aggregatedDataset.withColumn(
                 self.getRawPredictionCol(), rawPredictionUDF(aggregatedDataset[accColName]))

Review comment:
       ```suggestion
               aggregatedDataset = aggregatedDataset.withColumn(
                   self.getRawPredictionCol(), aggregatedDataset[accColName].cast(VectorUDT()))
   ```




-- 
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.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org