Skip to content

[SPARK-8200] [MLlib] Check for empty RDDs in StreamingLinearAlgorithm #6713

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
pparkkin wants to merge 5 commits into from
Closed

[SPARK-8200] [MLlib] Check for empty RDDs in StreamingLinearAlgorithm #6713

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
0bfc365
Test case for empty stream.
pparkkin Jun 9, 2015
393e36f
Ignore empty RDDs.
pparkkin Jun 9, 2015
54ad89e
Test case for empty stream.
pparkkin Jun 9, 2015
db234cf
Use !rdd.isEmpty.
pparkkin Jun 9, 2015
ff5cd78
Update strings to use interpolation.
pparkkin Jun 9, 2015
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 8 additions & 6 deletions mllib/src/main/scala/org/apache/spark/mllib/regression/StreamingLinearAlgorithm.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -83,13 +83,15 @@ abstract class StreamingLinearAlgorithm[
throw new IllegalArgumentException("Model must be initialized before starting training.")
}
data.foreachRDD { (rdd, time) =>
model = Some(algorithm.run(rdd, model.get.weights))
logInfo("Model updated at time %s".format(time.toString))
val display = model.get.weights.size match {
case x if x > 100 => model.get.weights.toArray.take(100).mkString("[", ",", "...")
case _ => model.get.weights.toArray.mkString("[", ",", "]")
if (!rdd.isEmpty) {
model = Some(algorithm.run(rdd, model.get.weights))
logInfo(s"Model updated at time ${time.toString}")
Copy link
Member

@srowen srowen Jun 10, 2015

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

(no need for toString here but don't worry about it)

val display = model.get.weights.size match {
case x if x > 100 => model.get.weights.toArray.take(100).mkString("[", ",", "...")
case _ => model.get.weights.toArray.mkString("[", ",", "]")
}
logInfo(s"Current model: weights, ${display}")
}
logInfo("Current model: weights, %s".format (display))
}
}

Expand Down
17 changes: 17 additions & 0 deletions ...c/test/scala/org/apache/spark/mllib/classification/StreamingLogisticRegressionSuite.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -158,4 +158,21 @@ class StreamingLogisticRegressionSuite extends SparkFunSuite with TestSuiteBase
val error = output.map(batch => batch.map(p => math.abs(p._1 - p._2)).sum / nPoints).toList
assert(error.head > 0.8 & error.last < 0.2)
}

// Test empty RDDs in a stream
test("handling empty RDDs in a stream") {
val model = new StreamingLogisticRegressionWithSGD()
.setInitialWeights(Vectors.dense(-0.1))
.setStepSize(0.01)
.setNumIterations(10)
val numBatches = 10
val emptyInput = Seq.empty[Seq[LabeledPoint]]
val ssc = setupStreams(emptyInput,
(inputDStream: DStream[LabeledPoint]) => {
model.trainOn(inputDStream)
model.predictOnValues(inputDStream.map(x => (x.label, x.features)))
}
)
val output: Seq[Seq[(Double, Double)]] = runStreams(ssc, numBatches, numBatches)
}
}
18 changes: 18 additions & 0 deletions mllib/src/test/scala/org/apache/spark/mllib/regression/StreamingLinearRegressionSuite.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -166,4 +166,22 @@ class StreamingLinearRegressionSuite extends SparkFunSuite with TestSuiteBase {
val error = output.map(batch => batch.map(p => math.abs(p._1 - p._2)).sum / nPoints).toList
assert((error.head - error.last) > 2)
}

// Test empty RDDs in a stream
test("handling empty RDDs in a stream") {
val model = new StreamingLinearRegressionWithSGD()
.setInitialWeights(Vectors.dense(0.0, 0.0))
.setStepSize(0.2)
.setNumIterations(25)
val numBatches = 10
val nPoints = 100
val emptyInput = Seq.empty[Seq[LabeledPoint]]
val ssc = setupStreams(emptyInput,
(inputDStream: DStream[LabeledPoint]) => {
model.trainOn(inputDStream)
model.predictOnValues(inputDStream.map(x => (x.label, x.features)))
}
)
val output: Seq[Seq[(Double, Double)]] = runStreams(ssc, numBatches, numBatches)
}
}