Github user srowen commented on a diff in the pull request:
https://github.com/apache/spark/pull/3637#discussion_r21527840
--- Diff: mllib/src/main/scala/org/apache/spark/ml/LabeledPoint.scala ---
@@ -0,0 +1,52 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.ml
+
+import scala.beans.BeanInfo
+
+import org.apache.spark.annotation.AlphaComponent
+import org.apache.spark.mllib.linalg.Vector
+
+/**
+ * :: AlphaComponent ::
+ * Class that represents an instance (data point) for prediction tasks.
+ *
+ * @param label Label to predict
+ * @param features List of features describing this instance
+ * @param weight Instance weight
+ */
+@AlphaComponent
+@BeanInfo
+case class LabeledPoint(label: Double, features: Vector, weight: Double) {
--- End diff --
I mean the former. Yeah, that's probably the downside. Each data element is
at least an object, and you can't have it reduce to a `double[]` under the hood.
In the second example, I think you'd only ever really want `MixedFeatures`
as an abstraction. There's no need to think of all `CategoricalFeatures` as a
special case deserving a unique abstraction.
I suppose if you abstract the entire training example as an object, and
allow accessors like `getNumericFeature(index: Int)`,
`getCategoricalFeature(index: Int)` you can still internally optimize the
storage while exposing a richer object representation. You get the type safety
and optimization opportunity.
Sure, an `Array[Double]` could easily be translated into one of the more
elaborate representations above. I suppose I myself wouldn't want to make it
_too_ easy to not think about types!
Anyway, up to your judgment really. There are arguments several ways here.
Worth a thought to see if the idea of a bit more abstraction appeals to you.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
