[GitHub] spark pull request #15428: [SPARK-17219][ML] enchanced NaN value handling in...
Github user srowen commented on a diff in the pull request: https://github.com/apache/spark/pull/15428#discussion_r83181476 --- Diff: mllib/src/main/scala/org/apache/spark/ml/feature/Bucketizer.scala --- @@ -73,15 +78,27 @@ final class Bucketizer @Since("1.4.0") (@Since("1.4.0") override val uid: String @Since("1.4.0") def setOutputCol(value: String): this.type = set(outputCol, value) + /** @group setParam */ + @Since("2.1.0") + def setHandleInvalid(value: String): this.type = set(handleInvalid, value) + setDefault(handleInvalid, "error") + @Since("2.0.0") override def transform(dataset: Dataset[_]): DataFrame = { transformSchema(dataset.schema) -val bucketizer = udf { feature: Double => - Bucketizer.binarySearchForBuckets($(splits), feature) +val bucketizer: UserDefinedFunction = udf { (feature: Double, flag: String) => + Bucketizer.binarySearchForBuckets($(splits), feature, flag) +} +val filteredDataset = { --- End diff -- I don't see that the method handles `NaN` below. What binarySearch returns is undefined. One place or the other I think this has to be explicitly handled. --- 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
[GitHub] spark pull request #15428: [SPARK-17219][ML] enchanced NaN value handling in...
Github user VinceShieh commented on a diff in the pull request: https://github.com/apache/spark/pull/15428#discussion_r82743072 --- Diff: mllib/src/main/scala/org/apache/spark/ml/feature/Bucketizer.scala --- @@ -73,15 +78,27 @@ final class Bucketizer @Since("1.4.0") (@Since("1.4.0") override val uid: String @Since("1.4.0") def setOutputCol(value: String): this.type = set(outputCol, value) + /** @group setParam */ + @Since("2.1.0") + def setHandleInvalid(value: String): this.type = set(handleInvalid, value) + setDefault(handleInvalid, "error") + @Since("2.0.0") override def transform(dataset: Dataset[_]): DataFrame = { transformSchema(dataset.schema) -val bucketizer = udf { feature: Double => - Bucketizer.binarySearchForBuckets($(splits), feature) +val bucketizer: UserDefinedFunction = udf { (feature: Double, flag: String) => + Bucketizer.binarySearchForBuckets($(splits), feature, flag) +} +val filteredDataset = { --- End diff -- Nope, actually, NaN will trigger an error later in binarySearchForBuckets as an invalid feature value if no special handling is made. --- 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
[GitHub] spark pull request #15428: [SPARK-17219][ML] enchanced NaN value handling in...
Github user srowen commented on a diff in the pull request: https://github.com/apache/spark/pull/15428#discussion_r82741203 --- Diff: mllib/src/main/scala/org/apache/spark/ml/param/shared/sharedParams.scala --- @@ -270,10 +270,10 @@ private[ml] trait HasFitIntercept extends Params { private[ml] trait HasHandleInvalid extends Params { /** - * Param for how to handle invalid entries. Options are skip (which will filter out rows with bad values), or error (which will throw an error). More options may be added later. + * Param for how to handle invalid entries. Options are skip (which will filter out rows with bad values), or error (which will throw an error), or keep (which will keep the bad values in certain way). More options may be added later. --- End diff -- I'm neutral on the complexity that this adds, but not against it. It gets a little funny to say "keep invalid data" but I think we discussed that on the JIRA --- 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
[GitHub] spark pull request #15428: [SPARK-17219][ML] enchanced NaN value handling in...
Github user srowen commented on a diff in the pull request: https://github.com/apache/spark/pull/15428#discussion_r82741770 --- Diff: mllib/src/main/scala/org/apache/spark/ml/feature/Bucketizer.scala --- @@ -128,8 +145,9 @@ object Bucketizer extends DefaultParamsReadable[Bucketizer] { * Binary searching in several buckets to place each data point. * @throws SparkException if a feature is < splits.head or > splits.last */ - private[feature] def binarySearchForBuckets(splits: Array[Double], feature: Double): Double = { -if (feature.isNaN) { + private[feature] def binarySearchForBuckets + (splits: Array[Double], feature: Double, flag: String): Double = { --- End diff -- Nit: I think the convention is to leave the open paren on the previous line Doesn't this need to handle "skip" and "error"? throw an exception on NaN if "error" or ignore it if "skip"? --- 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
[GitHub] spark pull request #15428: [SPARK-17219][ML] enchanced NaN value handling in...
Github user srowen commented on a diff in the pull request: https://github.com/apache/spark/pull/15428#discussion_r82741459 --- Diff: mllib/src/main/scala/org/apache/spark/ml/feature/Bucketizer.scala --- @@ -73,15 +78,27 @@ final class Bucketizer @Since("1.4.0") (@Since("1.4.0") override val uid: String @Since("1.4.0") def setOutputCol(value: String): this.type = set(outputCol, value) + /** @group setParam */ + @Since("2.1.0") + def setHandleInvalid(value: String): this.type = set(handleInvalid, value) + setDefault(handleInvalid, "error") + @Since("2.0.0") override def transform(dataset: Dataset[_]): DataFrame = { transformSchema(dataset.schema) -val bucketizer = udf { feature: Double => - Bucketizer.binarySearchForBuckets($(splits), feature) +val bucketizer: UserDefinedFunction = udf { (feature: Double, flag: String) => + Bucketizer.binarySearchForBuckets($(splits), feature, flag) +} +val filteredDataset = { --- End diff -- Doesn't this need to try to handle "error"? ``` val filteredDataSet = getHandleInvalid match { case "skip" => dataset.na.drop case "keep" => dataset case "error" => if (...dataset contains NaN...) { throw new IllegalArgumentException(...) } else { dataset } } ``` --- 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
[GitHub] spark pull request #15428: [SPARK-17219][ML] enchanced NaN value handling in...
GitHub user VinceShieh opened a pull request: https://github.com/apache/spark/pull/15428 [SPARK-17219][ML] enchanced NaN value handling in Bucketizer ## What changes were proposed in this pull request? This PR is an enhancement of PR with commit ID:57dc326bd00cf0a49da971e9c573c48ae28acaa2. NaN is a special type of value which is commonly seen as invalid. But We find that there are certain cases where NaN are also valuable, thus need special handling. We provided user when dealing NaN values with 3 options, to either reserve an extra bucket for NaN values, or remove the NaN values, or report an error, by passing "keep", "skip", or "error"(default) to setHandleInvalid. '''Before: val bucketizer: Bucketizer = new Bucketizer() .setInputCol("feature") .setOutputCol("result") .setSplits(splits) '''After: val bucketizer: Bucketizer = new Bucketizer() .setInputCol("feature") .setOutputCol("result") .setSplits(splits) .setHandleInvalid("keep") ## How was this patch tested? Tests added in QuantileDiscretizerSuite and BucketizerSuite Signed-off-by: VinceShiehYou can merge this pull request into a Git repository by running: $ git pull https://github.com/VinceShieh/spark spark-17219_followup Alternatively you can review and apply these changes as the patch at: https://github.com/apache/spark/pull/15428.patch To close this pull request, make a commit to your master/trunk branch with (at least) the following in the commit message: This closes #15428 commit a3e43086dcf6ecee20461567e2cc506db29f80a7 Author: VinceShieh Date: 2016-10-10T02:33:09Z [SPARK-17219][ML] enchance NaN value handling in Bucketizer This PR is an enhancement of PR with commit ID:57dc326bd00cf0a49da971e9c573c48ae28acaa2. We provided user when dealing NaN value in the dataset with 3 options, to either reserve an extra bucket for NaN values, or remove the NaN values, or report an error, by passing "keep", "skip", or "error"(default) to setHandleInvalid. '''Before: val bucketizer: Bucketizer = new Bucketizer() .setInputCol("feature") .setOutputCol("result") .setSplits(splits) '''After: val bucketizer: Bucketizer = new Bucketizer() .setInputCol("feature") .setOutputCol("result") .setSplits(splits) .setHandleInvalid("skip") Signed-off-by: VinceShieh --- 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