yunfengzhou-hub commented on a change in pull request #70:
URL: https://github.com/apache/flink-ml/pull/70#discussion_r834972977
##########
File path:
flink-ml-lib/src/main/java/org/apache/flink/ml/clustering/kmeans/KMeansModelParams.java
##########
@@ -21,27 +21,11 @@
import org.apache.flink.ml.common.param.HasDistanceMeasure;
import org.apache.flink.ml.common.param.HasFeaturesCol;
import org.apache.flink.ml.common.param.HasPredictionCol;
-import org.apache.flink.ml.param.IntParam;
-import org.apache.flink.ml.param.Param;
-import org.apache.flink.ml.param.ParamValidators;
/**
- * Params of {@link KMeansModel}.
+ * Params of {@link KMeansModel} and {@link OnlineKMeansModel}.
*
* @param <T> The class type of this instance.
*/
public interface KMeansModelParams<T>
- extends HasDistanceMeasure<T>, HasFeaturesCol<T>, HasPredictionCol<T> {
-
- Param<Integer> K =
Review comment:
In `KMeansModel`, `getK()` is never used, so if the input model data's K
is different from that of KMeansModel's K parameter, it would still function
correctly. Because of this I removed K from `KMeansModelParams`.
Now I agree that users may want to get to know the K of the model operator.
But in order to solve the concern above, I'd also like to check the model data
centroids length and the value of K parameter in `KMeansModel` and
`OnlineKMeansModel`, so as to make sure the K functions correctly.
--
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]
