On 3/28/13 12:03 PM, Thomas Neidhart wrote: > On 03/28/2013 07:46 PM, Thomas Neidhart wrote: >> On 03/28/2013 06:33 PM, Continuum@vmbuild wrote: >>> Group (shared) Maven 2 Build Definition (Java 1.5) >>> MIME-Version: 1.0 >>> Content-Type: text/plain; charset=us-ascii >>> Content-Transfer-Encoding: 7bit >>> X-Continuum-Build-Host: vmbuild >>> X-Continuum-Project-Id: 97 >>> X-Continuum-Project-Name: Commons Math >>> >>> Online report : >>> http://vmbuild.apache.org/continuum/buildResult.action?buildId=26645&projectId=97 >>> >>> Build statistics: >>> State: Failed >>> Previous State: Ok >>> Started at: Thu 28 Mar 2013 17:22:52 +0000 >>> Finished at: Thu 28 Mar 2013 17:33:23 +0000 >>> Total time: 10m 30s >>> Build Trigger: Schedule >>> Build Number: 1235 >>> Exit code: 1 >>> Building machine hostname: vmbuild >>> Operating system : Linux(unknown) >>> Java Home version : >>> java version "1.6.0_30" >>> Java(TM) SE Runtime Environment (build 1.6.0_30-b12) >>> Java HotSpot(TM) 64-Bit Server VM (build 20.5-b03, mixed mode) >>> >>> Builder version : >>> Apache Maven 2.2.1 (r801777; 2009-08-06 19:16:01+0000) >>> Java version: 1.6.0_30 >>> Java home: /usr/lib/jvm/jdk1.6.0_30/jre >>> Default locale: en_AU, platform encoding: UTF-8 >>> OS name: "linux" version: "2.6.32-41-server" arch: "amd64" >>> Family: "unix" >>> >>> **************************************************************************** >>> SCM Changes: >>> **************************************************************************** >>> Changed: psteitz @ Thu 28 Mar 2013 16:31:00 +0000 >>> Comment: Give Evan credit for recent patch and add him to contributors. >>> Files changed: >>> /commons/proper/math/trunk/pom.xml ( 1462206 ) >>> /commons/proper/math/trunk/src/changes/changes.xml ( 1462206 ) >>> >>> **************************************************************************** >>> Dependencies Changes: >>> **************************************************************************** >>> No dependencies changed >>> >>> >>> **************************************************************************** >>> Build Definition: >>> **************************************************************************** >>> POM filename: pom.xml >>> Goals: clean deploy >>> Arguments: --batch-mode -Pjava-1.5 -Dgpg.skip -Prelease >>> Build Fresh: false >>> Always Build: false >>> Default Build Definition: true >>> Schedule: COMMONS_SCHEDULE >>> Profile Name: Maven 2.2.1 >>> Description: Group (shared) Maven 2 Build Definition (Java 1.5) >>> >>> **************************************************************************** >>> Test Summary: >>> **************************************************************************** >>> Tests: 4974 >>> Failures: 1 >>> Errors: 0 >>> Success Rate: 99 >>> Total time: 549.5061 >>> >>> **************************************************************************** >>> Test Failures: >>> **************************************************************************** >>> >>> CorrelatedRandomVectorGeneratorTest >>> testSampleWithZeroCovariance : >>> java.lang.AssertionError >>> java.lang.AssertionError: Array comparison failure >>> Elements at index 1 differ. expected = 0.023006616 observed = >>> 0.024007487470287137 >>> at org.junit.Assert.fail(Assert.java:93) >>> at org.apache.commons.math3.TestUtils.assertEquals(TestUtils.java:365) >>> at org.apache.commons.math3.TestUtils.assertEquals(TestUtils.java:102) >>> at >>> org.apache.commons.math3.random.CorrelatedRandomVectorGeneratorTest.testSampler(CorrelatedRandomVectorGeneratorTest.java:193) >>> at >>> org.apache.commons.math3.random.CorrelatedRandomVectorGeneratorTest.testSampleWithZeroCovariance(CorrelatedRandomVectorGeneratorTest.java:167) >> fails only for specific seeds, e.g. with 628 as seed for the >> JDKRandomGenerator that is used in the test. > If I understand the test correctly, it assumes that the computed > covariance of 10000 sample vectors from the random generator equals the > given covariance matrix.
Yes, I think that is what the test is doing. > > Increasing the number of samples to 100000 seems to make the test more > robust. It is probably better to do what we have tried touniformly do elsewhere, which is to use a fixed seed for the underlying PRNG. I am curious, though, how we came up with the .001 and why we can get failures for some seeds at this level of sensitivity. I am not sure what an appropriate statistical test would be here. Might be good to file a bug or put a comment in the code so we don't forget to look into designing a statistical test of some kind. Phil > > Thomas > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > For additional commands, e-mail: dev-h...@commons.apache.org > > --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
