Skip to content
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.

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

Add default param grid for xgboost #175

Merged
merged 8 commits into from
Nov 7, 2018
Merged

Add default param grid for xgboost #175

merged 8 commits into from
Nov 7, 2018

Conversation

kinfaikan
Copy link
Contributor

Related issues
XGBoost does not have default param grid.

Describe the proposed solution
XGBoost has quite many hyperparameters. Consider only numRound, eta, maxDepth, and minChildWeight in default param grid.

Describe alternatives you've considered
NA

Additional context
NA

@kinfaikan kinfaikan changed the title add default param grid for xgboost Add default param grid for xgboost Nov 6, 2018
@tovbinm
Copy link
Collaborator

tovbinm commented Nov 6, 2018

@kinfaikan please add xgboost to model selector enums and tests

@@ -101,13 +101,15 @@ class BinaryClassificationModelSelectorTest extends FlatSpec with TestSparkConte

Spec(BinaryClassificationModelSelector.getClass) should "properly select models to try" in {
val modelSelector = BinaryClassificationModelSelector
.withCrossValidation(modelTypesToUse = Seq(BMT.OpLogisticRegression, BMT.OpRandomForestClassifier))
.setInput(label.asInstanceOf[Feature[RealNN]], features)
.withCrossValidation(
Copy link
Collaborator

Choose a reason for hiding this comment

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

@kinfaikan please mixin OpXGBoostQuietLogging trait into tests to avoid unnecessary logging.

Copy link
Collaborator

@tovbinm tovbinm left a comment

Choose a reason for hiding this comment

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

lgtm

@codecov
Copy link

codecov bot commented Nov 6, 2018

Codecov Report

Merging #175 into master will increase coverage by 17.35%.
The diff coverage is 100%.

Impacted file tree graph

@@             Coverage Diff             @@
##           master     #175       +/-   ##
===========================================
+ Coverage   68.93%   86.28%   +17.35%     
===========================================
  Files         305      305               
  Lines        9967     9977       +10     
  Branches      336      548      +212     
===========================================
+ Hits         6871     8609     +1738     
+ Misses       3096     1368     -1728
Impacted Files Coverage Δ
...p/stages/impl/selector/DefaultSelectorParams.scala 100% <100%> (ø) ⬆️
...ages/impl/regression/RegressionModelSelector.scala 95.23% <100%> (+0.5%) ⬆️
...sification/BinaryClassificationModelSelector.scala 96.15% <100%> (+0.5%) ⬆️
...ssification/MultiClassificationModelSelector.scala 95.23% <100%> (+68.92%) ⬆️
.../salesforce/op/features/FeatureBuilderMacros.scala 0% <0%> (-100%) ⬇️
.../op/features/types/FeatureTypeSparkConverter.scala 99.06% <0%> (+0.93%) ⬆️
...scala/com/salesforce/op/features/FeatureLike.scala 42.39% <0%> (+1.08%) ⬆️
...op/stages/impl/selector/ModelSelectorSummary.scala 91.3% <0%> (+1.08%) ⬆️
...rce/op/stages/impl/feature/PhoneNumberParser.scala 100% <0%> (+1.29%) ⬆️
...com/salesforce/op/evaluators/OpEvaluatorBase.scala 90.47% <0%> (+1.58%) ⬆️
... and 104 more

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 48314f7...cd65be7. Read the comment docs.

@tovbinm tovbinm merged commit b57b8b9 into master Nov 7, 2018
@tovbinm tovbinm deleted the kk/xgbParamGrid branch November 7, 2018 04:18
ericwayman pushed a commit that referenced this pull request Feb 8, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants