Hi , I have uploaded the patch to jira ,mahout-696-r4.patch is latest version
https://issues.apache.org/jira/browse/MAHOUT-696 You can use this command to run the trainAdaptiveLogisticRegression trainAdaptiveLogistic --input donut.csv --output d:\model --target color --categories 2 --predictors x shape --types numeric word --features 20 --passes 100 --scores --skipscorenum 99 And you can run TrainAdaptiveLogisticRegression under Eclipse using the following arguments: --input donut.csv --output d:\model --target color --categories 2 --predictors x shape --types numeric word --features 20 --passes 100 --scores --skipscorenum 99 On Mon, May 30, 2011 at 12:26 PM, Ted Dunning <ted.dunn...@gmail.com> wrote: > Do you have a test case that demonstrates this? > > On Sun, May 29, 2011 at 6:53 PM, Xiaobo Gu <guxiaobo1...@gmail.com> wrote: >> >> There is a internal buffer in AdaptiveLogisticRegression, the >> NullPointerException is caused when the backend crossfloderlearners >> starting training the examples. >> >> The default size of the buffer is 500, and the exception is caused >> when I put the 501'th example to ALR. >> >> >> On Mon, May 30, 2011 at 2:27 AM, Ted Dunning <ted.dunn...@gmail.com> >> wrote: >> > This usually means that you have fed the ALR enough data for it to push >> > a >> > batch of learning into the evolutionary algorithm. That means that >> > there >> > isn't any best result yet. >> > Getting that null doesn't impact the model, but you have to watch out >> > for >> > it. >> > >> > On Sun, May 29, 2011 at 1:23 AM, XiaoboGu <guxiaobo1...@gmail.com> >> > wrote: >> >> >> >> Hi, >> >> >> >> The main process for MAHOUT-696 is as following, but it will >> >> always >> >> cause a NullPointerException after the first call to getBest, can we >> >> continue training AdaptiveLogisticRegressions after using getBest() to >> >> score >> >> some new lines just as TrainLogistic does? >> >> >> >> >> >> >> >> double logPEstimate = 0; >> >> int k = 0; >> >> >> >> CsvRecordFactory csv = >> >> lmp.getCsvRecordFactory(); >> >> model = lmp.createAdaptiveLogisticRegression(); >> >> State<Wrapper, CrossFoldLearner> best = null; >> >> CrossFoldLearner learner = null; >> >> >> >> for (int pass = 0; pass < passes; pass++) { >> >> BufferedReader in = open(inputFile); >> >> >> >> // read variable names >> >> csv.firstLine(in.readLine()); >> >> >> >> String line = in.readLine(); >> >> >> >> while (line != null) { >> >> // for each new line, get target >> >> and predictors >> >> Vector input = new >> >> RandomAccessSparseVector(lmp.getNumFeatures()); >> >> int targetValue = >> >> csv.processLine(line, input); >> >> >> >> // update model >> >> model.train(targetValue, input); >> >> >> >> k ++; >> >> >> >> if (scores && (k % (skipscorenum >> >> + >> >> 1) == 0) ) { >> >> >> >> best = model.getBest(); >> >> if (null != best) { >> >> learner = >> >> best.getPayload().getLearner(); >> >> } >> >> if (learner != null) { >> >> // check performance >> >> while >> >> this is still news >> >> double logP = >> >> learner.logLikelihood(targetValue, input); >> >> if >> >> (!Double.isInfinite(logP)) { >> >> if (k < 20) { >> >> >> >> logPEstimate = (k * logPEstimate + logP) >> >> >> >> / (k + 1); >> >> } else { >> >> >> >> logPEstimate = 0.95 * logPEstimate + 0.05 >> >> >> >> * logP; >> >> } >> >> } >> >> double p = >> >> learner.classifyScalar(input); >> >> >> >> output.printf(Locale.ENGLISH, >> >> "%10d >> >> %2d >> >> %10.2f %2.4f %10.4f %10.4f\n", >> >> k, >> >> targetValue, >> >> >> >> learner.percentCorrect(), p, logP, >> >> >> >> logPEstimate); >> >> }else{ >> >> >> >> output.printf(Locale.ENGLISH, >> >> >> >> "%10d %2d %s\n", k, targetValue, >> >> >> >> "AdaptiveLogisticRegression is not ready for scoring ... "); >> >> } >> >> } >> >> >> >> >> >> line = in.readLine(); >> >> } >> >> in.close(); >> >> } >> >> >> >> >> >> >> >> >> >> 100 1 AdaptiveLogisticRegression is not ready for scoring ... >> >> 200 0 AdaptiveLogisticRegression is not ready for scoring ... >> >> 300 1 AdaptiveLogisticRegression is not ready for scoring ... >> >> 400 0 AdaptiveLogisticRegression is not ready for scoring ... >> >> 500 1 AdaptiveLogisticRegression is not ready for scoring ... >> >> Exception in thread "main" java.lang.IllegalStateException: >> >> java.lang.NullPointerException >> >> at >> >> >> >> org.apache.mahout.classifier.sgd.AdaptiveLogisticRegression.trainWithBufferedExamples(AdaptiveLogisticRegression.java:144) >> >> at >> >> >> >> org.apache.mahout.classifier.sgd.AdaptiveLogisticRegression.train(AdaptiveLogisticRegression.java:117) >> >> at >> >> >> >> org.apache.mahout.classifier.sgd.AdaptiveLogisticRegression.train(AdaptiveLogisticRegression.java:103) >> >> at >> >> >> >> org.apache.mahout.classifier.sgd.TrainAdaptiveLogistic.main(TrainAdaptiveLogistic.java:72) >> >> Caused by: java.lang.NullPointerException >> >> at >> >> >> >> org.apache.mahout.classifier.sgd.CrossFoldLearner.train(CrossFoldLearner.java:134) >> >> at >> >> >> >> org.apache.mahout.classifier.sgd.AdaptiveLogisticRegression$Wrapper.train(AdaptiveLogisticRegression.java:411) >> >> at >> >> >> >> org.apache.mahout.classifier.sgd.AdaptiveLogisticRegression$1.apply(AdaptiveLogisticRegression.java:128) >> >> at >> >> >> >> org.apache.mahout.classifier.sgd.AdaptiveLogisticRegression$1.apply(AdaptiveLogisticRegression.java:1) >> >> at >> >> >> >> org.apache.mahout.ep.EvolutionaryProcess$1.call(EvolutionaryProcess.java:146) >> >> at >> >> >> >> org.apache.mahout.ep.EvolutionaryProcess$1.call(EvolutionaryProcess.java:1) >> >> at >> >> java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303) >> >> at java.util.concurrent.FutureTask.run(FutureTask.java:138) >> >> at >> >> >> >> java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) >> >> at >> >> >> >> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) >> >> at java.lang.Thread.run(Thread.java:662) >> >> >> > >> > > >
