search

starling-lab / BoostSRL

BoostSRL: "Boosting for Statistical Relational Learning." A gradient-boosting based approach for learning different types of SRL models.
https://starling.utdallas.edu
GNU General Public License v3.0
32 stars 21 forks source link

No Boosting giving error #16

Open rodrigoazs opened 6 years ago

rodrigoazs commented 6 years ago

Hello,

When I set the -noBoost paramenter I am getting the following error:

Exception in thread "main" edu.wisc.cs.will.Utils.WILLthrownError: Probability was not set at edu.wisc.cs.will.Utils.Utils.error(Utils.java:484) at edu.wisc.cs.will.Boosting.RDN.RegressionRDNExample.getProbOfExample(RegressionRDNExample.java:176) at edu.wisc.cs.will.ILP.Regression.RegressionInfoHolderForRDN.populateExamples(RegressionInfoHolderForRDN.java:111) at edu.wisc.cs.will.ILP.SingleClauseNode.getRegressionInfoHolder(SingleClauseNode.java:1840) at edu.wisc.cs.will.ILP.SingleClauseNode.regressionFit(SingleClauseNode.java:1503) at edu.wisc.cs.will.ILP.SingleClauseNode.regressionFit(SingleClauseNode.java:1405) at edu.wisc.cs.will.ILP.ScoreRegressionNode.scoreThisNode(ScoreRegressionNode.java:35) at edu.wisc.cs.will.stdAIsearch.OpenList.insertByScoreIntoOpenList(OpenList.java:107) at edu.wisc.cs.will.stdAIsearch.OpenList.insertByScoreIntoOpenList(OpenList.java:216) at edu.wisc.cs.will.ILP.InitializeILPsearchSpace.initializeOpen(InitializeILPsearchSpace.java:187) at edu.wisc.cs.will.stdAIsearch.StateBasedSearchTask.initialize(StateBasedSearchTask.java:373) at edu.wisc.cs.will.stdAIsearch.StateBasedSearchTask.performSearchIteration(StateBasedSearchTask.java:505) at edu.wisc.cs.will.stdAIsearch.StateBasedSearchTask.performSearch(StateBasedSearchTask.java:440) at edu.wisc.cs.will.ILP.LearnOneClause.performSearch(LearnOneClause.java:1684) at edu.wisc.cs.will.ILP.LearnOneClause.performSearch(LearnOneClause.java:1708) at edu.wisc.cs.will.ILP.ILPouterLoop.executeOuterLoop(ILPouterLoop.java:686) at edu.wisc.cs.will.Boosting.RDN.LearnBoostedRDN.getWILLTree(LearnBoostedRDN.java:396) at edu.wisc.cs.will.Boosting.RDN.LearnBoostedRDN.learnRDN(LearnBoostedRDN.java:234) at edu.wisc.cs.will.Boosting.RDN.LearnBoostedRDN.learnNextModel(LearnBoostedRDN.java:129) at edu.wisc.cs.will.Boosting.RDN.RunBoostedRDN.learn(RunBoostedRDN.java:390) at edu.wisc.cs.will.Boosting.Common.RunBoostedModels.learnModel(RunBoostedModels.java:77) at edu.wisc.cs.will.Boosting.RDN.RunBoostedRDN.runJob(RunBoostedRDN.java:42) at edu.wisc.cs.will.Boosting.Common.RunBoostedModels.main(RunBoostedModels.java:220)

What am I doing wrong?

Thank you.

hayesall commented 6 years ago

Hey @rodrigoazs, -noBoost has been broken since before we began using git. I don't think it's anything you're doing wrong, fixing it is on the list of TODOs.

boost-starai commented 6 years ago

A suggestion is that you could instead run the code with numTrees set to 1 and maxTreeDepth to a larger number say 6-10. This will make sure that you still learn one regression tree.

We will fix this and push the code out in the next week. Thanks for bringing this to our attention.

Thanks Sriraam

On Apr 21, 2018, at 6:29 PM, Alexander L. Hayes notifications@github.com<mailto:notifications@github.com> wrote:

Hey @rodrigoazshttps://github.com/rodrigoazs, -noBoost has been broken since before we began using git. I don't think it's anything you're doing wrong, fixing it is on the list of TODOs.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/starling-lab/BoostSRL/issues/16#issuecomment-383341117, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AXip1bH9y4KC032yyfbkhvhMZY_tdI94ks5tq8DTgaJpZM4TeqCr.

rodrigoazs commented 6 years ago

I was about to ask if they were equivalent. Thank you.