### What changes were proposed in this pull request?
1, when `predictionCol` and `quantilesCol` are both set, we only need one prediction for each row: prediction is just the variable `lambda` in `predictQuantiles`;
2, in the computation of variable `quantiles` in `predictQuantiles`, a pre-computed vector `val baseQuantiles = $(quantileProbabilities).map(q => math.exp(math.log(-math.log1p(-q)) * scale))` can be reused for each row;
### Why are the changes needed?
avoid redundant computation in transform, like what we did in `ProbabilisticClassificationModel`, `GaussianMixtureModel`, etc
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
existing testsuite
Closes#30000 from zhengruifeng/aft_predict_transform_opt.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
Propagate LibSVM options to Hadoop configs in the LibSVM datasource.
### Why are the changes needed?
There is a bug that when running:
```scala
spark.read.format("libsvm").options(conf).load(path)
```
The underlying file system will not receive the `conf` options.
### Does this PR introduce _any_ user-facing change?
Yes. After the changes, for example, users should read files from Azure Data Lake successfully:
```scala
def hadoopConf1() = Map[String, String](
s"fs.adl.oauth2.access.token.provider.type" -> "ClientCredential",
s"fs.adl.oauth2.client.id" -> dbutils.secrets.get(scope = "...", key = "..."),
s"fs.adl.oauth2.credential" -> dbutils.secrets.get(scope = "...", key = "..."),
s"fs.adl.oauth2.refresh.url" -> s"https://login.microsoftonline.com/.../oauth2/token")
val df = spark.read.format("libsvm").options(hadoopConf1).load("adl://....azuredatalakestore.net/foldersp1/...")
```
and not get the following exception because the settings above are not propagated to the filesystem:
```java
java.lang.IllegalArgumentException: No value for fs.adl.oauth2.access.token.provider found in conf file.
at ....adl.AdlFileSystem.getNonEmptyVal(AdlFileSystem.java:820)
at ....adl.AdlFileSystem.getCustomAccessTokenProvider(AdlFileSystem.java:220)
at ....adl.AdlFileSystem.getAccessTokenProvider(AdlFileSystem.java:257)
at ....adl.AdlFileSystem.initialize(AdlFileSystem.java:164)
at org.apache.hadoop.fs.FileSystem.createFileSystem(FileSystem.java:2669)
```
### How was this patch tested?
Added UT to `LibSVMRelationSuite`.
Closes#29984 from MaxGekk/ml-option-propagation.
Authored-by: Max Gekk <max.gekk@gmail.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
RowMatrix contains a computation based on spark.driver.maxResultSize. However, when this value is set to 0, the computation fails (log of 0). The fix is simply to correctly handle this setting, which means unlimited result size, by using a tree depth of 1 in the RowMatrix method.
### Why are the changes needed?
Simple bug fix to make several Spark ML functions which use RowMatrix run correctly in this case.
### Does this PR introduce _any_ user-facing change?
Not other than the bug fix of course.
### How was this patch tested?
Existing RowMatrix tests plus a new test.
Closes#29925 from srowen/SPARK-33043.
Authored-by: Sean Owen <srowen@gmail.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
This PR proposes to promote the stability annotation to `Evolving` for MLEvent traits/classes.
### Why are the changes needed?
The feature is released in Spark 3.0.0 having SPARK-26818 as the last change in Feb. 2020, and haven't changed in Spark 3.0.1. (There's no change more than a half of year.)
While we'd better to wait for some minor releases to consider the API as stable, it would worth to promote to Evolving so that we clearly state that we support the API.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Just changed the annotation, no tests required.
Closes#29887 from HeartSaVioR/SPARK-33011.
Authored-by: Jungtaek Lim (HeartSaVioR) <kabhwan.opensource@gmail.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>
### What changes were proposed in this pull request?
The purpose of this pr is to resolve SPARK-32972, total of 51 Scala failed test cases and 3 Java failed test cases were fixed, the main change of this pr as follow:
- Specified `Seq` to `scala.collection.Seq` in case match `Seq` scene and `x.asInstanceOf[Seq[T]]` scene
- Use `Row.getSeq[T]` instead of `Row.getAs[Seq]`
- Manual call `toMap` method to convert `MapView` to `Map` in Scala 2.13
- Change the tol in the last test to 0.75 to pass `RandomForestRegressorSuite#training with sample weights` in Scala 2.13
### Why are the changes needed?
We need to support a Scala 2.13 build.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
- Scala 2.12: Pass the Jenkins or GitHub Action
- Scala 2.13: Pass GitHub 2.13 Build Action
Do the follow:
```
dev/change-scala-version.sh 2.13
mvn clean install -DskipTests -pl mllib -Pscala-2.13 -am
mvn test -pl mllib -Pscala-2.13 -fn
```
**Before**
```
[ERROR] Errors:
[ERROR] JavaVectorIndexerSuite.vectorIndexerAPI:51 » ClassCast scala.collection.conver...
[ERROR] JavaWord2VecSuite.testJavaWord2Vec:51 » Spark Job aborted due to stage failure...
[ERROR] JavaPrefixSpanSuite.runPrefixSpanSaveLoad:79 » Spark Job aborted due to stage ...
Tests: succeeded 1567, failed 51, canceled 0, ignored 7, pending 0
*** 51 TESTS FAILED ***
```
**After**
```
[INFO] Tests run: 122, Failures: 0, Errors: 0, Skipped: 0
Tests: succeeded 1617, failed 0, canceled 0, ignored 7, pending 0
All tests passed.
```
Closes#29857 from LuciferYang/fix-mllib-2.
Authored-by: yangjie01 <yangjie01@baidu.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
1, update the comment: `Note, the relevant columns must also be set in inputCols` -> `Note, the relevant columns should also be set in inputCols`;
2, add a check, and if there are `categoricalCols` not set in `inputCols`, log.warn it;
### Why are the changes needed?
1, there is no check to make sure `categoricalCols` are all set in `inputCols`, to keep existing behavior, update this comments;
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
repl
Closes#29868 from zhengruifeng/feature_hash_cat_doc.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
pre-compute the output indices of numerical columns, instead of computing them on each row.
### Why are the changes needed?
for a numerical column, its output index is a hash of its `col_name`, we can pre-compute it at first, instead of computing it on each row.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
existing testsuites
Closes#29850 from zhengruifeng/hash_opt.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: zhengruifeng <ruifengz@foxmail.com>
### What changes were proposed in this pull request?
`HashingTF` use `util.collection.OpenHashMap` instead of `mutable.HashMap`
### Why are the changes needed?
according to `util.collection.OpenHashMap` 's doc:
> This map is about 5X faster than java.util.HashMap, while using much less space overhead.
according to performance tests like ([Simple microbenchmarks comparing Scala vs Java mutable map performance ](https://gist.github.com/pchiusano/1423303)), `mutable.HashMap` maybe more inefficient than `java.util.HashMap`
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
existing testsuites
Closes#29852 from zhengruifeng/hashingtf_opt.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
revert blockify gmm
### Why are the changes needed?
WeichenXu123 and I thought we should use memory size instead of number of rows to blockify instance; then if a buffer's size is large and determined by number of rows, we should discard it.
In GMM, we found that the pre-allocated memory maybe too large and should be discarded:
```
transient private lazy val auxiliaryPDFMat = DenseMatrix.zeros(blockSize, numFeatures)
```
We had some offline discuss and thought it is better to revert blockify GMM.
### Does this PR introduce _any_ user-facing change?
blockSize added in master branch will be removed
### How was this patch tested?
existing testsuites
Closes#29782 from zhengruifeng/unblockify_gmm.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: zhengruifeng <ruifengz@foxmail.com>
### What changes were proposed in this pull request?
1, simplify the aggregation by get `count` via `summary.count`
2, ignore nan values like the old impl:
```
val relativeError = 0.05
val approxQuantile = numNearestNeighbors.toDouble / count + relativeError
val modelDatasetWithDist = modelDataset.withColumn(distCol, hashDistCol)
if (approxQuantile >= 1) {
modelDatasetWithDist
} else {
val hashThreshold = modelDatasetWithDist.stat
.approxQuantile(distCol, Array(approxQuantile), relativeError)
// Filter the dataset where the hash value is less than the threshold.
modelDatasetWithDist.filter(hashDistCol <= hashThreshold(0))
}
```
### Why are the changes needed?
simplify the aggregation
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
existing testsuites
Closes#29778 from zhengruifeng/lsh_nit.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: zhengruifeng <ruifengz@foxmail.com>
### What changes were proposed in this pull request?
This PR fix code which causes error when build with sbt and Scala 2.13 like as follows.
```
[error] [warn] /home/kou/work/oss/spark-scala-2.13/external/kafka-0-10/src/main/scala/org/apache/spark/streaming/kafka010/KafkaRDD.scala:251: method with a single empty parameter list overrides method without any parameter list
[error] [warn] override def hasNext(): Boolean = requestOffset < part.untilOffset
[error] [warn]
[error] [warn] /home/kou/work/oss/spark-scala-2.13/external/kafka-0-10/src/main/scala/org/apache/spark/streaming/kafka010/KafkaRDD.scala:294: method with a single empty parameter list overrides method without any parameter list
[error] [warn] override def hasNext(): Boolean = okNext
```
More specifically, what this PR fixes are
* Methods which has an empty parameter list and overrides an method which has no parameter list.
```
override def hasNext(): Boolean = okNext
```
* Methods which has no parameter list and overrides an method which has an empty parameter list.
```
override def next: (Int, Double) = {
```
* Infix operator expression that the operator wraps.
```
3L * math.min(k, numFeatures) * math.min(k, numFeatures)
3L * math.min(k, numFeatures) * math.min(k, numFeatures) +
+ math.max(math.max(k, numFeatures), 4L * math.min(k, numFeatures)
math.max(math.max(k, numFeatures), 4L * math.min(k, numFeatures) *
* math.min(k, numFeatures) + 4L * math.min(k, numFeatures))
```
### Why are the changes needed?
For building Spark with sbt and Scala 2.13.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
After this change and #29742 applied, compile passed with the following command.
```
build/sbt -Pscala-2.13 -Phive -Phive-thriftserver -Pyarn -Pkubernetes compile test:compile
```
Closes#29745 from sarutak/fix-code-for-sbt-and-spark-2.13.
Authored-by: Kousuke Saruta <sarutak@oss.nttdata.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>
### What changes were proposed in this pull request?
In the PR, I propose to fix an issue with LibSVM datasource when both of the following are true:
* no user specified schema
* some file paths contain escaped glob metacharacters, such as `[``]`, `{``}`, `*` etc.
The fix is based on another bug fix for CSV/JSON datasources https://github.com/apache/spark/pull/29659.
### Why are the changes needed?
To fix the issue when the follow two queries try to read from paths `[abc]`:
```scala
spark.read.format("libsvm").load("""/tmp/\[abc\].csv""").show
```
but would end up hitting an exception:
```
Path does not exist: file:/private/var/folders/p3/dfs6mf655d7fnjrsjvldh0tc0000gn/T/spark-6ef0ae5e-ff9f-4c4f-9ff4-0db3ee1f6a82/[abc]/part-00000-26406ab9-4e56-45fd-a25a-491c18a05e76-c000.libsvm;
org.apache.spark.sql.AnalysisException: Path does not exist: file:/private/var/folders/p3/dfs6mf655d7fnjrsjvldh0tc0000gn/T/spark-6ef0ae5e-ff9f-4c4f-9ff4-0db3ee1f6a82/[abc]/part-00000-26406ab9-4e56-45fd-a25a-491c18a05e76-c000.libsvm;
at org.apache.spark.sql.execution.datasources.DataSource$.$anonfun$checkAndGlobPathIfNecessary$3(DataSource.scala:770)
at org.apache.spark.util.ThreadUtils$.$anonfun$parmap$2(ThreadUtils.scala:373)
at scala.concurrent.Future$.$anonfun$apply$1(Future.scala:659)
at scala.util.Success.$anonfun$map$1(Try.scala:255)
at scala.util.Success.map(Try.scala:213)
```
### Does this PR introduce _any_ user-facing change?
Yes
### How was this patch tested?
Added UT to `LibSVMRelationSuite`.
Closes#29670 from MaxGekk/globbing-paths-when-inferring-schema-ml.
Authored-by: Max Gekk <max.gekk@gmail.com>
Signed-off-by: Wenchen Fan <wenchen@databricks.com>
### What changes were proposed in this pull request?
Fix double caching in KMeans/BiKMeans:
1, let the callers of `runWithWeight` to pass whether `handlePersistence` is needed;
2, persist and unpersist inside of `runWithWeight`;
3, persist the `norms` if needed according to the comments;
### Why are the changes needed?
avoid double caching
### Does this PR introduce _any_ user-facing change?
no
### How was this patch tested?
existing testsuites
Closes#29501 from zhengruifeng/kmeans_handlePersistence.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
The strict requirement for the vocabulary to remain non-empty has been removed in this pull request.
Link to the discussion: http://apache-spark-user-list.1001560.n3.nabble.com/Ability-to-have-CountVectorizerModel-vocab-as-empty-td38396.html
### Why are the changes needed?
This soothens running it across the corner cases. Without this, the user has to manupulate the data in genuine case, which may be a perfectly fine valid use-case.
Question: Should we a log when empty vocabulary is found instead?
### Does this PR introduce _any_ user-facing change?
May be a slight change. If someone has put a try-catch to detect an empty vocab. Then that behavior would no longer stand still.
### How was this patch tested?
1. Added testcase to `fit` generating an empty vocabulary
2. Added testcase to `transform` with empty vocabulary
Request to review: srowen hhbyyh
Closes#29482 from purijatin/spark_32662.
Authored-by: Jatin Puri <purijatin@gmail.com>
Signed-off-by: Huaxin Gao <huaxing@us.ibm.com>
### What changes were proposed in this pull request?
set params default values in trait Params for feature and tuning in both Scala and Python.
### Why are the changes needed?
Make ML has the same default param values between estimator and its corresponding transformer, and also between Scala and Python.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Existing and modified tests
Closes#29153 from huaxingao/default2.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: Huaxin Gao <huaxing@us.ibm.com>
### What changes were proposed in this pull request?
for binary `LogisticRegressionModel`:
1, keep variables `_threshold` and `_rawThreshold` instead of computing them on each instance;
2, in `raw2probabilityInPlace`, make use of the characteristic that the sum of probability is 1.0;
### Why are the changes needed?
for better performance
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
existing testsuite and performace test in REPL
Closes#29255 from zhengruifeng/pred_opt.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: Huaxin Gao <huaxing@us.ibm.com>
### What changes were proposed in this pull request?
Add training summary to MultilayerPerceptronClassificationModel...
### Why are the changes needed?
so that user can get the training process status, such as loss value of each iteration and total iteration number.
### Does this PR introduce _any_ user-facing change?
Yes
MultilayerPerceptronClassificationModel.summary
MultilayerPerceptronClassificationModel.evaluate
### How was this patch tested?
new tests
Closes#29250 from huaxingao/mlp_summary.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
logParam `thresholds` in DT/GBT/FM/LR/MLP
### Why are the changes needed?
param `thresholds` is logged in NB/RF, but not in other ProbabilisticClassifier
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
existing testsuites
Closes#29257 from zhengruifeng/instr.logParams_add_thresholds.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: Huaxin Gao <huaxing@us.ibm.com>
### What changes were proposed in this pull request?
Updates to scalatest 3.2.0. Though it looks large, it is 99% changes to the new location of scalatest classes.
### Why are the changes needed?
3.2.0+ has a fix that is required for Scala 2.13.3+ compatibility.
### Does this PR introduce _any_ user-facing change?
No, only affects tests.
### How was this patch tested?
Existing tests.
Closes#29196 from srowen/SPARK-32398.
Authored-by: Sean Owen <srowen@gmail.com>
Signed-off-by: Dongjoon Hyun <dongjoon@apache.org>
### What changes were proposed in this pull request?
use while-loop instead of the recursive way
### Why are the changes needed?
3% ~ 10% faster
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
existing testsuites
Closes#29095 from zhengruifeng/tree_pred_opt.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
set params default values in trait ...Params in both Scala and Python.
I will do this in two PRs. I will change classification, regression, clustering and fpm in this PR. Will change the rest in another PR.
### Why are the changes needed?
Make ML has the same default param values between estimator and its corresponding transformer, and also between Scala and Python.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Existing tests
Closes#29112 from huaxingao/set_default.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: Huaxin Gao <huaxing@us.ibm.com>
### What changes were proposed in this pull request?
Small improvement in the error message shown to user https://github.com/apache/spark/blob/master/mllib/src/main/scala/org/apache/spark/mllib/util/MLUtils.scala#L537-L538
### Why are the changes needed?
The current behavior is an exception is thrown but without any specific details on the cause
```
Caused by: java.lang.IllegalArgumentException: requirement failedCaused by: java.lang.IllegalArgumentException: requirement failed at scala.Predef$.require(Predef.scala:212) at org.apache.spark.mllib.util.MLUtils$.fastSquaredDistance(MLUtils.scala:508) at org.apache.spark.mllib.clustering.EuclideanDistanceMeasure$.fastSquaredDistance(DistanceMeasure.scala:232) at org.apache.spark.mllib.clustering.EuclideanDistanceMeasure.isCenterConverged(DistanceMeasure.scala:190) at org.apache.spark.mllib.clustering.KMeans$$anonfun$runAlgorithm$4.apply(KMeans.scala:336) at org.apache.spark.mllib.clustering.KMeans$$anonfun$runAlgorithm$4.apply(KMeans.scala:334) at scala.collection.MapLike$MappedValues$$anonfun$foreach$3.apply(MapLike.scala:245) at scala.collection.MapLike$MappedValues$$anonfun$foreach$3.apply(MapLike.scala:245) at scala.collection.TraversableLike$WithFilter$$anonfun$foreach$1.apply(TraversableLike.scala:733) at scala.collection.mutable.HashMap$$anonfun$foreach$1.apply(HashMap.scala:130) at scala.collection.mutable.HashMap$$anonfun$foreach$1.apply(HashMap.scala:130) at scala.collection.mutable.HashTable$class.foreachEntry(HashTable.scala:236) at scala.collection.mutable.HashMap.foreachEntry(HashMap.scala:40) at scala.collection.mutable.HashMap.foreach(HashMap.scala:130) at scala.collection.TraversableLike$WithFilter.foreach(TraversableLike.scala:732) at scala.collection.MapLike$MappedValues.foreach(MapLike.scala:245) at org.apache.spark.mllib.clustering.KMeans.runAlgorithm(KMeans.scala:334) at org.apache.spark.mllib.clustering.KMeans.run(KMeans.scala:251) at org.apache.spark.mllib.clustering.KMeans.run(KMeans.scala:233)
```
### Does this PR introduce _any_ user-facing change?
Yes, this PR adds an explanation message to be shown to user when requirement check is not meant
### How was this patch tested?
manually
Closes#29115 from dzlab/patch/SPARK-32315.
Authored-by: dzlab <dzlabs@outlook.com>
Signed-off-by: Huaxin Gao <huaxing@us.ibm.com>
### What changes were proposed in this pull request?
Same as https://github.com/apache/spark/pull/29078 and https://github.com/apache/spark/pull/28971 . This makes the rest of the default modules (i.e. those you get without specifying `-Pyarn` etc) compile under Scala 2.13. It does not close the JIRA, as a result. this also of course does not demonstrate that tests pass yet in 2.13.
Note, this does not fix the `repl` module; that's separate.
### Why are the changes needed?
Eventually, we need to support a Scala 2.13 build, perhaps in Spark 3.1.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Existing tests. (2.13 was not tested; this is about getting it to compile without breaking 2.12)
Closes#29111 from srowen/SPARK-29292.3.
Authored-by: Sean Owen <srowen@gmail.com>
Signed-off-by: Dongjoon Hyun <dongjoon@apache.org>
### What changes were proposed in this pull request?
Add training summary for FMClassificationModel...
### Why are the changes needed?
so that user can get the training process status, such as loss value of each iteration and total iteration number.
### Does this PR introduce _any_ user-facing change?
Yes
FMClassificationModel.summary
FMClassificationModel.evaluate
### How was this patch tested?
new tests
Closes#28960 from huaxingao/fm_summary.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: Huaxin Gao <huaxing@us.ibm.com>
# What changes were proposed in this pull request?
current problems:
```
mlp = MultilayerPerceptronClassifier(layers=[2, 2, 2], seed=123)
model = mlp.fit(df)
path = tempfile.mkdtemp()
model_path = path + "/mlp"
model.save(model_path)
model2 = MultilayerPerceptronClassificationModel.load(model_path)
self.assertEqual(model2.getSolver(), "l-bfgs") # this fails because model2.getSolver() returns 'auto'
model2.transform(df)
# this fails with Exception pyspark.sql.utils.IllegalArgumentException: MultilayerPerceptronClassifier_dec859ed24ec parameter solver given invalid value auto.
```
FMClassifier/Regression and GeneralizedLinearRegression have the same problems.
Here are the root cause of the problems:
1. In HasSolver, both Scala and Python default solver to 'auto'
2. On Scala side, mlp overrides the default of solver to 'l-bfgs', FMClassifier/Regression overrides the default of solver to 'adamW', and glr overrides the default of solver to 'irls'
3. On Scala side, mlp overrides the default of solver in MultilayerPerceptronClassificationParams, so both MultilayerPerceptronClassification and MultilayerPerceptronClassificationModel have 'l-bfgs' as default
4. On Python side, mlp overrides the default of solver in MultilayerPerceptronClassification, so it has default as 'l-bfgs', but MultilayerPerceptronClassificationModel doesn't override the default so it gets the default from HasSolver which is 'auto'. In theory, we don't care about the solver value or any other params values for MultilayerPerceptronClassificationModel, because we have the fitted model already. That's why on Python side, we never set default values for any of the XXXModel.
5. when calling getSolver on the loaded mlp model, it calls this line of code underneath:
```
def _transfer_params_from_java(self):
"""
Transforms the embedded params from the companion Java object.
"""
......
# SPARK-14931: Only check set params back to avoid default params mismatch.
if self._java_obj.isSet(java_param):
value = _java2py(sc, self._java_obj.getOrDefault(java_param))
self._set(**{param.name: value})
......
```
that's why model2.getSolver() returns 'auto'. The code doesn't get the default Scala value (in this case 'l-bfgs') to set to Python param, so it takes the default value (in this case 'auto') on Python side.
6. when calling model2.transform(df), it calls this underneath:
```
def _transfer_params_to_java(self):
"""
Transforms the embedded params to the companion Java object.
"""
......
if self.hasDefault(param):
pair = self._make_java_param_pair(param, self._defaultParamMap[param])
pair_defaults.append(pair)
......
```
Again, it gets the Python default solver which is 'auto', and this caused the Exception
7. Currently, on Scala side, for some of the algorithms, we set default values in the XXXParam, so both estimator and transformer get the default value. However, for some of the algorithms, we only set default in estimators, and the XXXModel doesn't get the default value. On Python side, we never set defaults for the XXXModel. This causes the default value inconsistency.
8. My proposed solution: set default params in XXXParam for both Scala and Python, so both the estimator and transformer have the same default value for both Scala and Python. I currently only changed solver in this PR. If everyone is OK with the fix, I will change all the other params as well.
I hope my explanation makes sense to your folks :)
### Why are the changes needed?
Fix bug
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
existing and new tests
Closes#29060 from huaxingao/solver_parity.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
1, GeneralizedLinearRegressionSummary compute several statistics on single pass
2, LinearRegressionSummary use metrics.count
### Why are the changes needed?
avoid extra passes on the dataset
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
existing testsuites
Closes#28990 from zhengruifeng/glr_summary_opt.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: Huaxin Gao <huaxing@us.ibm.com>
### What changes were proposed in this pull request?
Add summary to RandomForestClassificationModel...
### Why are the changes needed?
so user can get a summary of this classification model, and retrieve common metrics such as accuracy, weightedTruePositiveRate, roc (for binary), pr curves (for binary), etc.
### Does this PR introduce _any_ user-facing change?
Yes
```
RandomForestClassificationModel.summary
RandomForestClassificationModel.evaluate
```
### How was this patch tested?
Add new tests
Closes#28913 from huaxingao/rf_summary.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
Add training summary for LinearSVCModel......
### Why are the changes needed?
so that user can get the training process status, such as loss value of each iteration and total iteration number.
### Does this PR introduce _any_ user-facing change?
Yes
```LinearSVCModel.summary```
```LinearSVCModel.evaluate```
### How was this patch tested?
new tests
Closes#28884 from huaxingao/svc_summary.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
Adding support to Association Rules in Spark ml.fpm.
### Why are the changes needed?
Support is an indication of how frequently the itemset of an association rule appears in the database and suggests if the rules are generally applicable to the dateset. Refer to [wiki](https://en.wikipedia.org/wiki/Association_rule_learning#Support) for more details.
### Does this PR introduce _any_ user-facing change?
Yes. Associate Rules now have support measure
### How was this patch tested?
existing and new unit test
Closes#28903 from huaxingao/fpm.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
Add a generic ClassificationSummary trait
### Why are the changes needed?
Add a generic ClassificationSummary trait so all the classification models can use it to implement summary.
Currently in classification, we only have summary implemented in ```LogisticRegression```. There are requests to implement summary for ```LinearSVCModel``` in https://issues.apache.org/jira/browse/SPARK-20249 and to implement summary for ```RandomForestClassificationModel``` in https://issues.apache.org/jira/browse/SPARK-23631. If we add a generic ClassificationSummary trait and put all the common code there, we can easily add summary to ```LinearSVCModel``` and ```RandomForestClassificationModel```, and also add summary to all the other classification models.
We can use the same approach to add a generic RegressionSummary trait to regression package and implement summary for all the regression models.
### Does this PR introduce _any_ user-facing change?
### How was this patch tested?
existing tests
Closes#28710 from huaxingao/summary_trait.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
This patch adds user-specified fold column support to `CrossValidator`. User can assign fold numbers to dataset instead of letting Spark do random splits.
### Why are the changes needed?
This gives `CrossValidator` users more flexibility in splitting folds.
### Does this PR introduce _any_ user-facing change?
Yes, a new `foldCol` param is added to `CrossValidator`. User can use it to specify custom fold splitting.
### How was this patch tested?
Added unit tests.
Closes#28704 from viirya/SPARK-31777.
Authored-by: Liang-Chi Hsieh <viirya@gmail.com>
Signed-off-by: Liang-Chi Hsieh <liangchi@uber.com>
### What changes were proposed in this pull request?
In LogisticRegression and LinearRegression, if set maxIter=n, the model.summary.totalIterations returns n+1 if the training procedure does not drop out. This is because we use ```objectiveHistory.length``` as totalIterations, but ```objectiveHistory``` contains init sate, thus ```objectiveHistory.length``` is 1 larger than number of training iterations.
### Why are the changes needed?
correctness
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
add new tests and also modify existing tests
Closes#28786 from huaxingao/summary_iter.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
Add instance weight support in LinearRegressionSummary
### Why are the changes needed?
LinearRegression and RegressionMetrics support instance weight. We should support instance weight in LinearRegressionSummary too.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
add new test
Closes#28772 from huaxingao/lir_weight_summary.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
Add instance weight support in LogisticRegressionSummary
### Why are the changes needed?
LogisticRegression, MulticlassClassificationEvaluator and BinaryClassificationEvaluator support instance weight. We should support instance weight in LogisticRegressionSummary too.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Add new tests
Closes#28657 from huaxingao/weighted_summary.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
In the algorithms that support instance weight, add checks to make sure instance weight is not negative.
### Why are the changes needed?
instance weight has to be >= 0.0
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Manually tested
Closes#28621 from huaxingao/weight_check.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
Add weight support in ClusteringEvaluator
### Why are the changes needed?
Currently, BinaryClassificationEvaluator, RegressionEvaluator, and MulticlassClassificationEvaluator support instance weight, but ClusteringEvaluator doesn't, so we will add instance weight support in ClusteringEvaluator.
### Does this PR introduce _any_ user-facing change?
Yes.
ClusteringEvaluator.setWeightCol
### How was this patch tested?
add new unit test
Closes#28553 from huaxingao/weight_evaluator.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
add getMetrics in Evaluators to get the corresponding Metrics instance, so users can use it to get any of the metrics scores. For example:
```
val trainer = new LinearRegression
val model = trainer.fit(dataset)
val predictions = model.transform(dataset)
val evaluator = new RegressionEvaluator()
val metrics = evaluator.getMetrics(predictions)
val rmse = metrics.rootMeanSquaredError
val r2 = metrics.r2
val mae = metrics.meanAbsoluteError
val variance = metrics.explainedVariance
```
### Why are the changes needed?
Currently, Evaluator.evaluate only access to one metrics, but most users may need to get multiple metrics. This PR adds getMetrics in all the Evaluators, so users can use it to get an instance of the corresponding Metrics to get any of the metrics they want.
### Does this PR introduce _any_ user-facing change?
Yes. Add getMetrics in Evaluators.
For example:
```
/**
* Get a RegressionMetrics, which can be used to get any of the regression
* metrics such as rootMeanSquaredError, meanSquaredError, etc.
*
* param dataset a dataset that contains labels/observations and predictions.
* return RegressionMetrics
*/
Since("3.1.0")
def getMetrics(dataset: Dataset[_]): RegressionMetrics
```
### How was this patch tested?
Add new unit tests
Closes#28590 from huaxingao/getMetrics.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
In QuantileDiscretizer.getDistinctSplits, before invoking distinct, normalize all -0.0 and 0.0 to be 0.0
```
for (i <- 0 until splits.length) {
if (splits(i) == -0.0) {
splits(i) = 0.0
}
}
```
### Why are the changes needed?
Fix bug.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Unit test.
#### Manually test:
~~~scala
import scala.util.Random
val rng = new Random(3)
val a1 = Array.tabulate(200)(_=>rng.nextDouble * 2.0 - 1.0) ++ Array.fill(20)(0.0) ++ Array.fill(20)(-0.0)
import spark.implicits._
val df1 = sc.parallelize(a1, 2).toDF("id")
import org.apache.spark.ml.feature.QuantileDiscretizer
val qd = new QuantileDiscretizer().setInputCol("id").setOutputCol("out").setNumBuckets(200).setRelativeError(0.0)
val model = qd.fit(df1) // will raise error in spark master.
~~~
### Explain
scala `0.0 == -0.0` is True but `0.0.hashCode == -0.0.hashCode()` is False. This break the contract between equals() and hashCode() If two objects are equal, then they must have the same hash code.
And array.distinct will rely on elem.hashCode so it leads to this error.
Test code on distinct
```
import scala.util.Random
val rng = new Random(3)
val a1 = Array.tabulate(200)(_=>rng.nextDouble * 2.0 - 1.0) ++ Array.fill(20)(0.0) ++ Array.fill(20)(-0.0)
a1.distinct.sorted.foreach(x => print(x.toString + "\n"))
```
Then you will see output like:
```
...
-0.009292684662246975
-0.0033280686465135823
-0.0
0.0
0.0022219556032221366
0.02217419561977274
...
```
Closes#28498 from WeichenXu123/SPARK-31676.
Authored-by: Weichen Xu <weichen.xu@databricks.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
Expose hashFunc property in HashingTF
Some third-party library such as mleap need to access it.
See background description here:
https://github.com/combust/mleap/pull/665#issuecomment-621258623
### Why are the changes needed?
See https://github.com/combust/mleap/pull/665#issuecomment-621258623
### Does this PR introduce any user-facing change?
No. Only add a package private constructor.
### How was this patch tested?
N/A
Closes#28413 from WeichenXu123/hashing_tf_expose_hashfunc.
Authored-by: Weichen Xu <weichen.xu@databricks.com>
Signed-off-by: Xiangrui Meng <meng@databricks.com>
### What changes were proposed in this pull request?
1, add new param blockSize;
2, if blockSize==1, keep original behavior, code path trainOnRows;
3, if blockSize>1, standardize and stack input vectors to blocks (like ALS/MLP), code path trainOnBlocks
### Why are the changes needed?
performance gain on dense dataset HIGGS:
1, save about 45% RAM;
2, 3X faster with openBLAS
### Does this PR introduce any user-facing change?
add a new expert param `blockSize`
### How was this patch tested?
added testsuites
Closes#27473 from zhengruifeng/blockify_gmm.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: zhengruifeng <ruifengz@foxmail.com>
### What changes were proposed in this pull request?
When input column lengths can not be inferred and handleInvalid = "keep", VectorAssembler will throw a runtime exception. However the error message with this exception is not consistent. I change the content of this error message to make it work properly.
### Why are the changes needed?
This is a bug. Here is a simple example to reproduce it.
```
// create a df without vector size
val df = Seq(
(Vectors.dense(1.0), Vectors.dense(2.0))
).toDF("n1", "n2")
// only set vector size hint for n1 column
val hintedDf = new VectorSizeHint()
.setInputCol("n1")
.setSize(1)
.transform(df)
// assemble n1, n2
val output = new VectorAssembler()
.setInputCols(Array("n1", "n2"))
.setOutputCol("features")
.setHandleInvalid("keep")
.transform(hintedDf)
// because only n1 has vector size, the error message should tell us to set vector size for n2 too
output.show()
```
Expected error message:
```
Can not infer column lengths with handleInvalid = "keep". Consider using VectorSizeHint to add metadata for columns: [n2].
```
Actual error message:
```
Can not infer column lengths with handleInvalid = "keep". Consider using VectorSizeHint to add metadata for columns: [n1, n2].
```
This introduce difficulties when I try to resolve this exception, for I do not know which column required vectorSizeHint. This is especially troublesome when you have a large number of columns to deal with.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Add test in VectorAssemblerSuite.
Closes#28487 from fan31415/SPARK-31671.
Lead-authored-by: fan31415 <fan12356789@gmail.com>
Co-authored-by: yijiefan <fanyije@gmail.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
1, add new param blockSize;
2, add a new class InstanceBlock;
3, if blockSize==1, keep original behavior; if blockSize>1, stack input vectors to blocks (like ALS/MLP);
4, if blockSize>1, standardize the input outside of optimization procedure;
### Why are the changes needed?
it will obtain performance gain on dense datasets, such as epsilon
1, reduce RAM to persist traing dataset; (save about 40% RAM)
2, use Level-2 BLAS routines; (~10X speedup)
### Does this PR introduce _any_ user-facing change?
Yes, a new param is added
### How was this patch tested?
existing and added testsuites
Closes#28473 from zhengruifeng/blockify_aft.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: zhengruifeng <ruifengz@foxmail.com>
### What changes were proposed in this pull request?
Add ANOVASelector and FValueSelector to PySpark
### Why are the changes needed?
ANOVASelector and FValueSelector have been implemented in Scala. We need to implement these in Python as well.
### Does this PR introduce _any_ user-facing change?
Yes. Add Python version of ANOVASelector and FValueSelector
### How was this patch tested?
new doctest
Closes#28464 from huaxingao/selector_py.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: zhengruifeng <ruifengz@foxmail.com>
### What changes were proposed in this pull request?
1, add new param blockSize;
2, add a new class InstanceBlock;
3, if blockSize==1, keep original behavior; if blockSize>1, stack input vectors to blocks (like ALS/MLP);
4, if blockSize>1, standardize the input outside of optimization procedure;
### Why are the changes needed?
it will obtain performance gain on dense datasets, such as `epsilon`
1, reduce RAM to persist traing dataset; (save about 40% RAM)
2, use Level-2 BLAS routines; (up to 6X(squaredError)~12X(huber) speedup)
### Does this PR introduce _any_ user-facing change?
Yes, a new param is added
### How was this patch tested?
existing and added testsuites
Closes#28471 from zhengruifeng/blockify_lir_II.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: zhengruifeng <ruifengz@foxmail.com>
### What changes were proposed in this pull request?
1, reorg the `fit` method in LR to several blocks (`createModel`, `createBounds`, `createOptimizer`, `createInitCoefWithInterceptMatrix`);
2, add new param blockSize;
3, if blockSize==1, keep original behavior, code path `trainOnRows`;
4, if blockSize>1, standardize and stack input vectors to blocks (like ALS/MLP), code path `trainOnBlocks`
### Why are the changes needed?
On dense dataset `epsilon_normalized.t`:
1, reduce RAM to persist traing dataset; (save about 40% RAM)
2, use Level-2 BLAS routines; (4x ~ 5x faster)
### Does this PR introduce _any_ user-facing change?
Yes, a new param is added
### How was this patch tested?
existing and added testsuites
Closes#28458 from zhengruifeng/blockify_lor_II.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: zhengruifeng <ruifengz@foxmail.com>
### What changes were proposed in this pull request?
Implement abstract Selector. Put the common code among ```ANOVASelector```, ```ChiSqSelector```, ```FValueSelector``` and ```VarianceThresholdSelector``` to Selector.
### Why are the changes needed?
code reuse
### Does this PR introduce any user-facing change?
No
### How was this patch tested?
Existing tests
Closes#27978 from huaxingao/spark-31127.
Authored-by: Huaxin Gao <huaxing@us.ibm.com>
Signed-off-by: zhengruifeng <ruifengz@foxmail.com>
### What changes were proposed in this pull request?
1, add new param `blockSize`;
2, add a new class InstanceBlock;
3, **if `blockSize==1`, keep original behavior; if `blockSize>1`, stack input vectors to blocks (like ALS/MLP);**
4, if `blockSize>1`, standardize the input outside of optimization procedure;
### Why are the changes needed?
1, reduce RAM to persist traing dataset; (save about 40% RAM)
2, use Level-2 BLAS routines; (4x ~ 5x faster on dataset `epsilon`)
### Does this PR introduce any user-facing change?
Yes, a new param is added
### How was this patch tested?
existing and added testsuites
Closes#28349 from zhengruifeng/blockify_svc_II.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: zhengruifeng <ruifengz@foxmail.com>
### What changes were proposed in this pull request?
1, make AFT reuse common functions in `ml.optim`, rather than making its own impl.
### Why are the changes needed?
The logic in optimizing AFT is quite similar to other algorithms like other algs based on `RDDLossFunction`,
We should reuse the common functions.
### Does this PR introduce any user-facing change?
No
### How was this patch tested?
existing testsuites
Closes#28404 from zhengruifeng/mv_aft_optim.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
What changes were proposed in this pull request?
1.Add class info output in org.apache.spark.ml.util.SchemaUtils#checkColumnType to distinct Vectors in ml and mllib
2.Add unit test
Why are the changes needed?
the catalogString doesn't distinguish Vectors in ml and mllib when mllib vector misused in ml
https://issues.apache.org/jira/browse/SPARK-31400
Does this PR introduce any user-facing change?
No
How was this patch tested?
Unit test is added
Closes#28347 from TJX2014/master-catalogString-distinguish-Vectors-in-ml-and-mllib.
Authored-by: TJX2014 <xiaoxingstack@gmail.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
apply Lemma 1 in [Using the Triangle Inequality to Accelerate K-Means](https://www.aaai.org/Papers/ICML/2003/ICML03-022.pdf):
> Let x be a point, and let b and c be centers. If d(b,c)>=2d(x,b) then d(x,c) >= d(x,b);
It can be directly applied in EuclideanDistance, but not in CosineDistance.
However, for CosineDistance we can luckily get a variant in the space of radian/angle.
### Why are the changes needed?
It help improving the performance of prediction and training (mostly)
### Does this PR introduce any user-facing change?
No
### How was this patch tested?
existing testsuites
Closes#27758 from zhengruifeng/km_triangle.
Authored-by: zhengruifeng <ruifengz@foxmail.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
