[
https://issues.apache.org/jira/browse/FLINK-3919?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15307443#comment-15307443
]
ASF GitHub Bot commented on FLINK-3919:
---------------------------------------
Github user chobeat commented on a diff in the pull request:
https://github.com/apache/flink/pull/1996#discussion_r65142014
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/math/distributed/DistributedRowMatrix.scala
---
@@ -0,0 +1,167 @@
+/*
+ * 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.flink.ml.math.distributed
+
+import java.lang
+
+import breeze.linalg.{CSCMatrix => BreezeSparseMatrix, Matrix =>
BreezeMatrix, Vector => BreezeVector}
+import org.apache.flink.api.common.functions.RichGroupReduceFunction
+import org.apache.flink.api.common.typeinfo.TypeInformation
+import org.apache.flink.api.scala._
+import org.apache.flink.ml.math.{Matrix => FlinkMatrix, _}
+import org.apache.flink.util.Collector
+import org.apache.flink.ml.math.Breeze._
+import scala.collection.JavaConversions._
+
+/**
+ *
+ * @param numRowsOpt If None, will be calculated from the DataSet.
+ * @param numColsOpt If None, will be calculated from the DataSet.
+ */
+class DistributedRowMatrix(data: DataSet[IndexedRow],
+ numRowsOpt: Option[Int] = None,
+ numColsOpt: Option[Int] = None)
+ extends DistributedMatrix {
+
+ lazy val getNumRows: Int = numRowsOpt match {
+ case Some(rows) => rows
+ case None => data.count().toInt
+ }
+
+ lazy val getNumCols: Int = numColsOpt match {
+ case Some(cols) => cols
+ case None => calcCols
+ }
--- End diff --
I've found a flaw in this way of thought. It was already in my code but we
are keeping doing the same error: the matrix representation is sparse so
counting the actual rows is wrong. We should count the max index provided. If
no info is given about the matrix dimensionality, we assume that the row with
the highest indices is the last at the bottom of the matrix. Counting them
would give a totally wrong result.
> Distributed Linear Algebra: row-based matrix
> --------------------------------------------
>
> Key: FLINK-3919
> URL: https://issues.apache.org/jira/browse/FLINK-3919
> Project: Flink
> Issue Type: New Feature
> Components: Machine Learning Library
> Reporter: Simone Robutti
> Assignee: Simone Robutti
>
> Distributed matrix implementation as a DataSet of IndexedRow and related
> operations
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
