This is an automated email from the ASF dual-hosted git repository.
weichenxu123 pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/spark.git
The following commit(s) were added to refs/heads/master by this push:
new 6a89f654bb54 [SPARK-52130][FOLLOW-UP][ML][CONNECT] Refine error
message when model.summary is evicted
6a89f654bb54 is described below
commit 6a89f654bb54de73f20f07c10ca6f7ff186bbeb6
Author: Weichen Xu <weichen...@databricks.com>
AuthorDate: Fri May 30 12:51:47 2025 +0800
[SPARK-52130][FOLLOW-UP][ML][CONNECT] Refine error message when
model.summary is evicted
### What changes were proposed in this pull request?
Refine error message when model.summary is evicted.
In Spark Connect ML handler, `model.summary.XXX` invocation has different
code path than `model.evaluate`, and it should raise readable error message
when model.summary is evicted, otherwise the MLHandler will raise
NullPointerException which is not friendly to end-users.
### Why are the changes needed?
Refine error message.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Manually.
### Was this patch authored or co-authored using generative AI tooling?
No.
Closes #51048 from WeichenXu123/improve-scml-summary-error.
Authored-by: Weichen Xu <weichen...@databricks.com>
Signed-off-by: Weichen Xu <weichen...@databricks.com>
---
.../org/apache/spark/sql/connect/ml/MLHandler.scala | 17 ++++++++++++-----
1 file changed, 12 insertions(+), 5 deletions(-)
diff --git
a/sql/connect/server/src/main/scala/org/apache/spark/sql/connect/ml/MLHandler.scala
b/sql/connect/server/src/main/scala/org/apache/spark/sql/connect/ml/MLHandler.scala
index 927f959f714b..d40b70ba0813 100644
---
a/sql/connect/server/src/main/scala/org/apache/spark/sql/connect/ml/MLHandler.scala
+++
b/sql/connect/server/src/main/scala/org/apache/spark/sql/connect/ml/MLHandler.scala
@@ -28,7 +28,7 @@ import org.apache.spark.ml.{Estimator, EstimatorUtils, Model,
Transformer}
import org.apache.spark.ml.evaluation.Evaluator
import org.apache.spark.ml.param.{ParamMap, Params}
import org.apache.spark.ml.tree.TreeConfig
-import org.apache.spark.ml.util.{MLWritable, Summary}
+import org.apache.spark.ml.util.{HasTrainingSummary, MLWritable, Summary}
import org.apache.spark.sql.DataFrame
import org.apache.spark.sql.connect.common.LiteralValueProtoConverter
import org.apache.spark.sql.connect.ml.Serializer.deserializeMethodArguments
@@ -223,10 +223,17 @@ private[connect] object MLHandler extends Logging {
.build()
case proto.MlCommand.CommandCase.FETCH =>
- val helper = AttributeHelper(
- sessionHolder,
- mlCommand.getFetch.getObjRef.getId,
- mlCommand.getFetch.getMethodsList.asScala.toArray)
+ val objRefId = mlCommand.getFetch.getObjRef.getId
+ val methods = mlCommand.getFetch.getMethodsList.asScala.toArray
+ val obj = sessionHolder.mlCache.get(objRefId)
+ if (obj != null && obj.isInstanceOf[HasTrainingSummary[_]]
+ && methods(0).getMethod == "summary"
+ && !obj.asInstanceOf[HasTrainingSummary[_]].hasSummary) {
+ throw MLCacheInvalidException(
+ objRefId,
+ sessionHolder.mlCache.getOffloadingTimeoutMinute)
+ }
+ val helper = AttributeHelper(sessionHolder, objRefId, methods)
val attrResult = helper.getAttribute
attrResult match {
case s: Summary =>
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@spark.apache.org
For additional commands, e-mail: commits-h...@spark.apache.org
