ELK problem with Protege 5.0.0 beta15

[cer...@gmail.com]

Hi I am trying to use the ELK 0.4.1  reasoner in Protege  5.0.0 beta15.  Running on a Windows 7 machine with service pack 1.

Ontology is https://code.google.com/p/cell-ontology/source/browse/trunk/src/ontology/cl-edit.owl
I don't actually seem to be getting reasoning to occur.   What can I do to get ELK to work?


When I try to run the reasoner I get the following errors
Error logged
org.semanticweb.owlapi.reasoner.ReasonerInternalException: org.semanticweb.elk.o
wl.exceptions.ElkRuntimeException: Could not start workers required for reasoner
 computation!
        at org.semanticweb.elk.owlapi.ElkExceptionConverter.convert(ElkException
Converter.java:80)
        at org.semanticweb.elk.owlapi.ElkConverter.convert(ElkConverter.java:119
)
        at org.semanticweb.elk.owlapi.ElkReasoner.isConsistent(ElkReasoner.java:
855)
        at org.protege.editor.owl.model.inference.OWLReasonerManagerImpl.getReas
onerStatus(OWLReasonerManagerImpl.java:236)
        at org.protege.editor.owl.model.inference.OWLReasonerManagerImpl$1.uncau
ghtException(OWLReasonerManagerImpl.java:277)
        at java.lang.Thread.dispatchUncaughtException(Unknown Source)
Caused by: org.semanticweb.elk.owl.exceptions.ElkRuntimeException: Could not sta
rt workers required for reasoner computation!
        at org.semanticweb.elk.reasoner.ReasonerComputation.process(ReasonerComp
utation.java:93)
        at org.semanticweb.elk.reasoner.consistency.ConsistencyChecking.process(
ConsistencyChecking.java:195)
        at org.semanticweb.elk.reasoner.stages.ConsistencyCheckingStage.executeS
tage(ConsistencyCheckingStage.java:64)
        at org.semanticweb.elk.reasoner.stages.AbstractReasonerStage.execute(Abs
tractReasonerStage.java:184)
        at org.semanticweb.elk.reasoner.stages.LoggingStageExecutor.execute(Logg
ingStageExecutor.java:53)
        at org.semanticweb.elk.reasoner.stages.AbstractStageExecutor.complete(Ab
stractStageExecutor.java:53)
        at org.semanticweb.elk.reasoner.stages.AbstractReasonerState.isInconsist
ent(AbstractReasonerState.java:341)
        at org.semanticweb.elk.owlapi.ElkReasoner.isConsistent(ElkReasoner.java:
849)
        ... 3 more

Thanks,
Ceri

[Yevgeny Kazakov]

Hi Ceri,

Could you check if the nightly build works for you:

https://code.google.com/p/elk-reasoner/wiki/GettingElk

I think, we had some problem with Protege 5 and we might have fixed it
in the unreleased version.

Best regards,

Yevgeny

> --
> You received this message because you are subscribed to the Google Groups
> "elk-reasoner-discussion" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to elk-reasoner-disc...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

[cer...@gmail.com]

Tried that, got farther before failing
        ...done
loading OntologyID(Anonymous-0) from file:/C:/Users/Ceri/Desktop/Cell%20OWL%20ed
it%20version/ontology/cl-edit.owl
loading OntologyID(OntologyIRI(<http://purl.obolibrary.org/obo/cl/imports/clo_im
port.owl>)) from file:/C:/Users/Ceri/Desktop/Cell%20OWL%20edit%20version/ontolog
y/imports/clo_import.owl
loading OntologyID(OntologyIRI(<http://purl.obolibrary.org/obo/cl/imports/go_imp
ort.owl>)) from http://purl.obolibrary.org/obo/cl/imports/go_import.owl
loading OntologyID(OntologyIRI(<http://purl.obolibrary.org/obo/cl/imports/pr_imp
ort.owl>)) from http://purl.obolibrary.org/obo/cl/imports/pr_import.owl
loading OntologyID(OntologyIRI(<http://purl.obolibrary.org/obo/cl/imports/pato_i
mport.owl>)) from http://purl.obolibrary.org/obo/cl/imports/pato_import.owl
loading OntologyID(OntologyIRI(<http://purl.obolibrary.org/obo/cl/imports/uberon
_import.owl>)) from http://purl.obolibrary.org/obo/cl/imports/uberon_import.owl
loading OntologyID(OntologyIRI(<http://purl.obolibrary.org/obo/cl/imports/chebi_
import.owl>)) from http://purl.obolibrary.org/obo/cl/imports/chebi_import.owl
Setting active ontology to OntologyID(OntologyIRI(<http://purl.obolibrary.org/ob
o/cl.owl>))
Rebuilding entity indices...
... rebuilt in 192 ms
... active ontology changed
Setting active ontology to OntologyID(OntologyIRI(<http://purl.obolibrary.org/ob
o/cl.owl>))
Rebuilding entity indices...
... rebuilt in 41 ms
... active ontology changed
ELK reasoner was created
[reasoner.indexing.IndexedObjectComplementOf]ELK does not support negative occur
rences of ObjectComplementOf. Reasoning might be incomplete!
[reasoner.indexing.IndexedObjectUnionOf]ELK does not support positive occurrence
s of ObjectUnionOf. Reasoning might be incomplete!
[reasoner.indexing.IndexedObjectUnionOf]ELK does not support positive occurrence
s of ObjectUnionOf. Reasoning might be incomplete!
[reasoner.indexing.IndexedObjectUnionOf]ELK does not support positive occurrence
s of ObjectUnionOf. Reasoning might be incomplete!
[reasoner.indexing.axiomIgnored]ELK does not support InverseObjectProperties. Ax
iom ignored:
InverseObjectProperties(<http://purl.obolibrary.org/obo/BFO_0000050> <http://pur
l.obolibrary.org/obo/BFO_0000051>)
[reasoner.indexing.IndexedObjectUnionOf]ELK does not support positive occurrence
s of ObjectUnionOf. Reasoning might be incomplete!
[reasoner.indexing.axiomIgnored]ELK does not support InverseObjectProperties. Ax
iom ignored:
InverseObjectProperties(<http://purl.obolibrary.org/obo/RO_0002203> <http://purl
.obolibrary.org/obo/RO_0002202>)
[reasoner.indexing.IndexedObjectUnionOf]ELK does not support positive occurrence
s of ObjectUnionOf. Reasoning might be incomplete!
[reasoner.indexing.IndexedObjectUnionOf]ELK does not support positive occurrence
s of ObjectUnionOf. Reasoning might be incomplete!
[reasoner.indexing.IndexedObjectUnionOf]ELK does not support positive occurrence
s of ObjectUnionOf. Reasoning might be incomplete!
[reasoner.indexing.IndexedObjectUnionOf]ELK does not support positive occurrence
s of ObjectUnionOf. Reasoning might be incomplete!
[reasoner.indexing.IndexedObjectUnionOf]ELK does not support positive occurrence
s of ObjectUnionOf. Reasoning might be incomplete!
[reasoner.indexing.IndexedObjectUnionOf]ELK does not support positive occurrence
s of ObjectUnionOf. Reasoning might be incomplete!
[reasoner.indexing.IndexedObjectUnionOf]ELK does not support positive occurrence
s of ObjectUnionOf. Reasoning might be incomplete!
[reasoner.indexing.IndexedObjectUnionOf]ELK does not support positive occurrence
s of ObjectUnionOf. Reasoning might be incomplete!
[reasoner.indexing.IndexedObjectUnionOf]ELK does not support positive occurrence
s of ObjectUnionOf. Reasoning might be incomplete!
[reasoner.indexing.IndexedObjectUnionOf]ELK does not support positive occurrence
s of ObjectUnionOf. Reasoning might be incomplete!
[reasoner.indexing.axiomIgnored]ELK does not support InverseObjectProperties. Ax
iom ignored:
InverseObjectProperties(<http://purl.obolibrary.org/obo/BFO_0000050> <http://pur
l.obolibrary.org/obo/BFO_0000051>)
Initializing the reasoner by performing the following steps:
        class hierarchy
        object property hierarchy
        data property hierarchy
        class assertions
        object property assertions
        same individuals
Loading of Axioms started
Loading of Axioms using 8 workers
Loading of Axioms finished in 2 ms
Property Saturation Initialization started
Property Saturation Initialization using 8 workers
Property Saturation Initialization finished in 1 ms
Reflexive Property Computation started
Reflexive Property Computation using 8 workers
Reflexive Property Computation finished in 11 ms
Error logged
org.semanticweb.elk.util.concurrent.computation.ComputationRuntimeException: Unc
aught exception in a worker thread:
        at org.semanticweb.elk.util.concurrent.computation.ComputationExecutor$W
orker.run(ComputationExecutor.java:188)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(Unknown Source)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(Unknown Source)
        at java.lang.Thread.run(Unknown Source)
Caused by: java.lang.RuntimeException: java.lang.InterruptedException
        at org.apache.felix.framework.BundleWiringImpl$BundleClassLoader.findCla
ss(BundleWiringImpl.java:2102)
        at org.apache.felix.framework.BundleWiringImpl.findClassOrResourceByDele
gation(BundleWiringImpl.java:1501)
        at org.apache.felix.framework.BundleWiringImpl.access$400(BundleWiringIm
pl.java:75)
        at org.apache.felix.framework.BundleWiringImpl$BundleClassLoader.loadCla
ss(BundleWiringImpl.java:1955)
        at java.lang.ClassLoader.loadClass(Unknown Source)
        at org.semanticweb.elk.reasoner.saturation.properties.ReflexivePropertyC
omputationFactory.getEngine(ReflexivePropertyComputationFactory.java:61)
        at org.semanticweb.elk.reasoner.saturation.properties.ReflexivePropertyC
omputationFactory.getEngine(ReflexivePropertyComputationFactory.java:44)
        at org.semanticweb.elk.util.concurrent.computation.ConcurrentComputation
WithInputs$Worker.run(ConcurrentComputationWithInputs.java:151)
        at org.semanticweb.elk.util.concurrent.computation.ComputationExecutor$W
orker.run(ComputationExecutor.java:186)
        ... 3 more
Caused by: java.lang.InterruptedException
        at java.lang.Object.wait(Native Method)
        at java.lang.Object.wait(Object.java:503)
        at org.apache.felix.framework.BundleWiringImpl$BundleClassLoader.findCla
ss(BundleWiringImpl.java:2097)
        ... 11 more
Property Saturation Initialization started
Property Saturation Initialization using 8 workers
Property Saturation Initialization finished in 0 ms
Reflexive Property Computation started
Reflexive Property Computation using 8 workers

Could not start workers required for reasoner computation!

Reflexive Property Computation finished in 0 ms
ELK reasoner has shut down
Protege terminated reasoner.

Error logged
org.semanticweb.owlapi.reasoner.ReasonerInternalException: org.semanticweb.elk.o
wl.exceptions.ElkRuntimeException: Could not start workers required for reasoner
 computation!
        at org.semanticweb.elk.owlapi.ElkExceptionConverter.convert(ElkException
Converter.java:80)
        at org.semanticweb.elk.owlapi.ElkConverter.convert(ElkConverter.java:119
)
        at org.semanticweb.elk.owlapi.ElkReasoner.isConsistent(ElkReasoner.java:

839)

        at org.protege.editor.owl.model.inference.OWLReasonerManagerImpl.getReas
onerStatus(OWLReasonerManagerImpl.java:236)
        at org.protege.editor.owl.model.inference.OWLReasonerManagerImpl$1.uncau
ghtException(OWLReasonerManagerImpl.java:277)
        at java.lang.Thread.dispatchUncaughtException(Unknown Source)
Caused by: org.semanticweb.elk.owl.exceptions.ElkRuntimeException: Could not sta
rt workers required for reasoner computation!

        at org.semanticweb.elk.reasoner.ReasonerComputationWithInputs.process(Re
asonerComputationWithInputs.java:94)
        at org.semanticweb.elk.reasoner.stages.PropertyReflexivityComputationSta
ge.executeStage(PropertyReflexivityComputationStage.java:58)
        at org.semanticweb.elk.reasoner.stages.AbstractReasonerStage.execute(Abs
tractReasonerStage.java:188)
        at org.semanticweb.elk.reasoner.stages.PropertyReflexivityComputationSta
ge.execute(PropertyReflexivityComputationStage.java:28)
        at org.semanticweb.elk.reasoner.stages.LoggingStageExecutor.execute(Logg
ingStageExecutor.java:54)
        at org.semanticweb.elk.reasoner.stages.AbstractStageExecutor.complete(Ab
stractStageExecutor.java:53)
        at org.semanticweb.elk.reasoner.stages.AbstractStageExecutor.complete(Ab
stractStageExecutor.java:47)
        at org.semanticweb.elk.reasoner.stages.AbstractStageExecutor.complete(Ab
stractStageExecutor.java:47)
        at org.semanticweb.elk.reasoner.stages.AbstractStageExecutor.complete(Ab
stractStageExecutor.java:47)
        at org.semanticweb.elk.reasoner.stages.AbstractReasonerState.isInconsist
ent(AbstractReasonerState.java:363)
        at org.semanticweb.elk.owlapi.ElkReasoner.isConsistent(ElkReasoner.java:
833)
        ... 3 more

> email to elk-reasoner-discussion+unsub...@googlegroups.com.

[Pavel Klinov]

Ceri,

Thanks for the report, we will try to reproduce this issue.

Cheers,
Pavel

>> > email to elk-reasoner-disc...@googlegroups.com.

>> > For more options, visit https://groups.google.com/d/optout.
>

> --
> You received this message because you are subscribed to the Google Groups
> "elk-reasoner-discussion" group.
> To unsubscribe from this group and stop receiving emails from it, send an

> email to elk-reasoner-disc...@googlegroups.com.

[Chris Mungall]

I get the same error as Ceri using the 0.4.1 distro with Protege 5: "Could not start workers required for reasoner computation". (starting Protege from the command line via "run.sh" on OS 10.8)

I switched to the nightly build, and it seemed at first that the problem remained. However, I persevered, and restarted the reasoner, and everything was hunky dory after that. Rather curious, but all good now.

Maybe time to make a new Elk release? What are the differences I can expect, if any, using the 0.5.0 nightly build, compared to 0.4.1? Changes.txt is silent on the matter.

[David Osumi-Sutherland]

On 6 Jan 2015, at 20:56, Chris Mungall <cjmu...@lbl.gov> wrote:

I get the same error as Ceri using the 0.4.1 distro with Protege 5: "Could not start workers required for reasoner computation". (starting Protege from the command line via "run.sh" on OS 10.8)

I switched to the nightly build, and it seemed at first that the problem remained. However, I persevered, and restarted the reasoner, and everything was hunky dory after that. Rather curious, but all good now.

Maybe time to make a new Elk release? What are the differences I can expect, if any, using the 0.5.0 nightly build, compared to 0.4.1? Changes.txt is silent on the matter.

Latest nightly build + GO editor's file freezes + Protege 5 - Protege freezes follwoing successful classification.  All fine with 0.4.1 in Protege 4.

Cheers,

David

[Yevgeny Kazakov]

Dear Cris, David, Ceri,

I think Pavel looked at this problem. See the ticket:

https://code.google.com/p/elk-reasoner/issues/detail?id=27

If I am not mistaken, the problem is caused by some incompatibilities
between Protege 4 and Protege 5 manifest files (they require different
versions of Apache Felix or something like that). Interestingly, ELK
worked fine with some previous betas of Protege 5.

Are you still using Protege 5 beta 15 or do you build Protege 5 from
sources? Beta 15 was released quite a while ago and there have been
some changes since then. I recall Pavel mentioned that the problem
does not take place anymore in the current master, but I might be
wrong -- the ticket is still open for some reason.

David, is your "GO editor's file" the same as "go-plus-deve.owl" from here:
http://geneontology.org/page/download-ontology#go-plus-dev.owl

I tried to load it in Protege but obtained a file not found exception
for some of the import dependencies:

java.io.FileNotFoundException:
http://geneontology.org/ontology/extensions/x-cell.owl

Also, your system (OS X 10.8) seems a bit dated. Can your try the same
on the latest OS X 10.10? Which version of Java do you have installed?

Regarding the release 0.5.0 of ELK, I cannot give you any estimates at
the moment.
The key feature of the new version is a native debugging support
(showing how subsumptions have been derived from axioms). See our
paper:
http://elk.semanticweb.org/publications/Kazakov-Klinov-EL-Tracing-ISWC2014.pdf

To use the debugger in Protege one would require an additional Protege
plugin, which is not provided in the nightly builds yet.

Best regards,

Yevgeny

[Pavel Klinov]

On Wed, Jan 7, 2015 at 4:16 PM, Yevgeny Kazakov
<yevgeny...@uni-ulm.de> wrote:
> Dear Cris, David, Ceri,
>
> I think Pavel looked at this problem. See the ticket:
>
> https://code.google.com/p/elk-reasoner/issues/detail?id=27
>
> If I am not mistaken, the problem is caused by some incompatibilities
> between Protege 4 and Protege 5 manifest files (they require different
> versions of Apache Felix or something like that). Interestingly, ELK
> worked fine with some previous betas of Protege 5.
>
> Are you still using Protege 5 beta 15 or do you build Protege 5 from
> sources? Beta 15 was released quite a while ago and there have been
> some changes since then. I recall Pavel mentioned that the problem
> does not take place anymore in the current master,

That's correct. I believe the problem has been resolved for Ceri.

> but I might be
> wrong -- the ticket is still open for some reason.

Most likely I wanted to keep it open until the fix is in master and
then forgot to close. I'll close it but we'll re-open if David's issue
turns out to be the same (and thus unfixed).

Cheers,
Pavel

[Chris Mungall]

On 7 Jan 2015, at 7:16, Yevgeny Kazakov wrote:

> Dear Cris, David, Ceri,
>
> I think Pavel looked at this problem. See the ticket:
>
> https://code.google.com/p/elk-reasoner/issues/detail?id=27

Confirmed, thanks

> If I am not mistaken, the problem is caused by some incompatibilities
> between Protege 4 and Protege 5 manifest files (they require different
> versions of Apache Felix or something like that). Interestingly, ELK
> worked fine with some previous betas of Protege 5.
>
> Are you still using Protege 5 beta 15 or do you build Protege 5 from
> sources?

OK, good suggestion. I generally use the pre-built version so I am in
sync with other users.

> Beta 15 was released quite a while ago and there have been
> some changes since then. I recall Pavel mentioned that the problem
> does not take place anymore in the current master, but I might be
> wrong -- the ticket is still open for some reason.
>
> David, is your "GO editor's file" the same as "go-plus-deve.owl" from
> here:
> http://geneontology.org/page/download-ontology#go-plus-dev.owl

Apologies, I will remove this file from the page. go-plus-dev is no
longer very relevant as most of the axioms have been rolled into
go-plus.

In general, members of the OWL community should use go-plus, available
from here:
http://geneontology.org/page/download-ontology#Currently_available

>
> I tried to load it in Protege but obtained a file not found exception
> for some of the import dependencies:
>
> java.io.FileNotFoundException:
> http://geneontology.org/ontology/extensions/x-cell.owl

Fixed if you use go-plus

> Also, your system (OS X 10.8) seems a bit dated. Can your try the same
> on the latest OS X 10.10? Which version of Java do you have installed?
>
> Regarding the release 0.5.0 of ELK, I cannot give you any estimates at
> the moment.

OK. We will direct people using Protege5 to use the nightly build for
now

> The key feature of the new version is a native debugging support
> (showing how subsumptions have been derived from axioms). See our
> paper:
> http://elk.semanticweb.org/publications/Kazakov-Klinov-EL-Tracing-ISWC2014.pdf
> To use the debugger in Protege one would require an additional Protege
> plugin, which is not provided in the nightly builds yet.

Great, thanks

[Yevgeny Kazakov]

David,

are you still experiencing the freezing problem?
I just tried loading and classifying go-plus.owl (version 2015-01.07) from
http://purl.obolibrary.org/obo/go/extensions/go-plus.owl
with the daily snapshot of ELK from yesterday (there were no changes
in this year so it should be the same as the one you tested)
ELK finished classification in about 5 seconds:

Protege did not freeze for although switching between tabs was a bit sluggish.
I have Mac OS X 10.10.1, java version "1.7.0_51"

The only thing I did with Protege 5 beta 15 is set -Xmx1G in run.sh
(with 512M it runs out of memory when loading).

Best regards,

Yevgeny


On Wed, Jan 7, 2015 at 11:04 AM, David Osumi-Sutherland
<dj...@gen.cam.ac.uk> wrote: