Github user manishamde commented on a diff in the pull request:
https://github.com/apache/spark/pull/79#discussion_r11103459
--- Diff:
mllib/src/main/scala/org/apache/spark/mllib/tree/impurity/Impurity.scala ---
@@ -0,0 +1,42 @@
+/*
+ * 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.mllib.tree.impurity
+
+/**
+ * Trail for calculating information gain
+ */
+trait Impurity extends Serializable {
+
+ /**
+ * information calculation for binary classification
+ * @param c0 count of instances with label 0
+ * @param c1 count of instances with label 1
+ * @return information value
+ */
+ def calculate(c0 : Double, c1 : Double): Double
+
+ /**
+ * information calculation for regression
+ * @param count number of instances
+ * @param sum sum of labels
+ * @param sumSquares summation of squares of the labels
+ * @return information value
+ */
+ def calculate(count: Double, sum: Double, sumSquares: Double): Double
--- End diff --
@mengxr The generic interface you noted is correct. However, I think
implementing this generic interface and the corresponding implementations is
not a minor code change. There are some assumptions in the bin aggregation code
that may need to be updated and it also requires adding partition-wise impurity
calculation and aggregation.
@mateiz As @mengxr noted, it's highly unlikely that a user will write their
own ```Impurity``` implementation. It's mostly an internal API and could be
addressed soon in a different PR.
I think we all agree (please correct me if I am wrong) the ```Impurity```
update belongs to a different PR. I can spend time on it immediately after this
PR is accepted.
Is this the correct method of marking a method as unstable using the
javadoc?
```<span class="badge" style="float: right; background-color:
darkblue;">ALPHA COMPONENT</span>```
---
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 [email protected] or file a JIRA ticket
with INFRA.
---
