flink-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From thvasilo <...@git.apache.org>
Subject [GitHub] flink pull request #2542: [FLINK-4613] Extend ALS to handle implicit feedbac...
Date Wed, 28 Sep 2016 08:08:45 GMT
Github user thvasilo commented on a diff in the pull request:

    https://github.com/apache/flink/pull/2542#discussion_r80862241
  
    --- Diff: flink-libraries/flink-ml/src/test/scala/org/apache/flink/ml/recommendation/ImplicitALSTest.scala
---
    @@ -0,0 +1,171 @@
    +/*
    + * 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.recommendation
    +
    +import org.apache.flink.ml.util.FlinkTestBase
    +import org.scalatest._
    +
    +import scala.language.postfixOps
    +import org.apache.flink.api.scala._
    +import org.apache.flink.core.testutils.CommonTestUtils
    +
    +class ImplicitALSTest
    +  extends FlatSpec
    +    with Matchers
    +    with FlinkTestBase {
    +
    +  override val parallelism = 2
    +
    +  behavior of "The modification of the alternating least squares (ALS) implementation"
+
    +    "for implicit feedback datasets."
    +
    +  it should "properly compute Y^T * Y, and factorize matrix" in {
    +    import ExampleMatrix._
    +
    +    val rand = scala.util.Random
    +    val numBlocks = 3
    +    // randomly split matrix to blocks
    +    val blocksY = Y
    +      // add a random block id to every row
    +      .map { row =>
    +        (rand.nextInt(numBlocks), row)
    +      }
    +      // get the block via grouping
    +      .groupBy(_._1).values
    +      // add a block id (-1) to each block
    +      .map(b => (-1, b.map(_._2)))
    +      .toSeq
    +
    +    // use Flink to compute YtY
    +    val env = ExecutionEnvironment.getExecutionEnvironment
    +
    +    val distribBlocksY = env.fromCollection(blocksY)
    +
    +    val YtY = ALS
    +      .computeXtX(distribBlocksY, factors)
    +      .collect().head
    +
    +    // check YtY size
    +    YtY.length should be (factors * (factors - 1) / 2 + factors)
    +
    +    // check result is as expected
    +    expectedUpperTriangleYtY
    +      .zip(YtY)
    +      .foreach { case (expected, result) =>
    +        result should be (expected +- 0.1)
    +      }
    +
    +    // temporary directory to avoid too few memory segments
    +    val tempDir = CommonTestUtils.getTempDir + "/"
    +
    +    // factorize matrix with implicit ALS
    +    val als = ALS()
    +      .setIterations(iterations)
    +      .setLambda(lambda)
    +      .setBlocks(blocks)
    +      .setNumFactors(factors)
    +      .setImplicit(true)
    +      .setAlpha(alpha)
    +      .setSeed(seed)
    +      .setTemporaryPath(tempDir)
    +
    +    val inputDS = env.fromCollection(implicitRatings)
    +
    +    als.fit(inputDS)
    +
    +    // check predictions on some user-item pairs
    +    val testData = env.fromCollection(expectedResult.map{
    +      case (userID, itemID, rating) => (userID, itemID)
    +    })
    +
    +    val predictions = als.predict(testData).collect()
    +
    +    predictions.length should equal(expectedResult.length)
    +
    +    val resultMap = expectedResult map {
    +      case (uID, iID, value) => (uID, iID) -> value
    +    } toMap
    +
    +    predictions foreach {
    +      case (uID, iID, value) => {
    +        resultMap.isDefinedAt((uID, iID)) should be(true)
    +
    +        value should be(resultMap((uID, iID)) +- 1e-5)
    +      }
    +    }
    +
    +  }
    +
    +}
    +
    +object ExampleMatrix {
    --- End diff --
    
    Data should go to the `Recommendation.scala` file, as with the plain ALS matrix.


---
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 infrastructure@apache.org or file a JIRA ticket
with INFRA.
---

Mime
View raw message