Guilhem Vuillier created SPARK-44585:
----------------------------------------
Summary: Fix warning condition in MLLib RankingMetrics ndcgAk
Key: SPARK-44585
URL: https://issues.apache.org/jira/browse/SPARK-44585
Project: Spark
Issue Type: Bug
Components: MLlib
Affects Versions: 3.4.1
Reporter: Guilhem Vuillier
The implementation of nDCG evaluation in MLLib with relevance score (added in
3.4.0, see https://issues.apache.org/jira/browse/SPARK-39446 and [pull
request|https://github.com/apache/spark/pull/36843]) implements the following
warning when the input data isn't correct: "# of ground truth set and # of
relevance value set should be equal, check input data"
The logic for raising warnings is faulty at the moment: it raises a warning
when the following conditions are both true:
# {{rel}} is empty
# {{lab.size}} and {{rel.size}} are not equal.
With the current logic, RankingMetrics will:
* raise incorrect warning when a user is using it in the "binary" mode (i.e.
no relevance values in the input)
* not raise warning (that could be necessary) when the user is using it in the
"non-binary" model (i.e. with relevance values in the input)
The logic should be to raise a warning should be:
# {{rel}} is *not empty*
# {{lab.size}} and {{rel.size}} are not equal.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
