[SPARK-29232][ML] Update the parameter maps of the DecisionTreeRegression/Classification Models #26154
Conversation
…sion/Classification Models
| } else { | ||
| GradientBoostedTrees.run(trainDataset, boostingStrategy, $(seed), $(featureSubsetStrategy)) | ||
| } | ||
| baseLearners.map(tree => copyValues(tree)) |
There was a problem hiding this comment.
maybe baseLearners.foreach(copyValues) ?
There was a problem hiding this comment.
Thanks! I will change to baseLearners.foreach(copyValues(_)) because copyValues takes two arguments.
| val trees = RandomForest | ||
| .run(instances, strategy, getNumTrees, getFeatureSubsetStrategy, getSeed, Some(instr)) | ||
| .map(_.asInstanceOf[DecisionTreeClassificationModel]) | ||
| trees.map(tree => copyValues(tree)) |
There was a problem hiding this comment.
trees.foreach(copyValues)
| val trees = RandomForest | ||
| .run(instances, strategy, getNumTrees, getFeatureSubsetStrategy, getSeed, Some(instr)) | ||
| .map(_.asInstanceOf[DecisionTreeRegressionModel]) | ||
| trees.map(tree => copyValues(tree)) |
|
One more question: Is this satisfied? |
|
Test build #112244 has finished for PR 26154 at commit
|
The trees in |
|
Test build #112248 has finished for PR 26154 at commit
|
|
Merged to master, thanks @huaxingao |
|
Thanks a lot! @zhengruifeng |
4 participants
What changes were proposed in this pull request?
The trees (Array[
DecisionTreeRegressionModel]) inRandomForestRegressionModelonly contains the default parameter value. Need to update the parameter maps for these trees.Same issues in
RandomForestClassifier,GBTClassifierandGBTRegressorWhy are the changes needed?
User wants to access each individual tree and build the trees back up for the random forest estimator. This doesn't work because trees don't have the correct parameter values
Does this PR introduce any user-facing change?
Yes. Now the trees in
RandomForestRegressionModel,RandomForestClassifier,GBTClassifierandGBTRegressorhave the correct parameter values.How was this patch tested?
Add tests