[OTDev] How to upload a dataset ...
Tobias Girschick tobias.girschick at in.tum.deMon Sep 6 08:48:09 CEST 2010
- Previous message: [OTDev] How to upload a dataset ...
- Next message: [OTDev] How to upload a dataset ...
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
On the first glance I'd say, you're not adding the data entries to the
dataset after creating them. Try adding a line like this one:
dataset.addProperty(OT.dataEntry, dataEntry);
regards,
Tobias
On Fri, 2010-09-03 at 23:28 +0530, surajit ray wrote:
> Hi,
>
> Lemme take a look into that ....
>
> heres my code creating the dataset ... can you tell me whats wrong with it ?
>
> *****************************************
> this.resultDataset = createModel();
> Individual dataset =
> resultDataset.createIndividual(resultDataset.getOntClass(OT.OTClass.Dataset.getNS()));
> OT.OTClass.Dataset.createOntClass(resultDataset);
> OT.OTClass.DataEntry.createOntClass(resultDataset);
> OT.OTClass.Feature.createOntClass(resultDataset);
> OT.OTClass.FeatureValue.createOntClass(resultDataset);
> OT.OTClass.Compound.createOntClass(resultDataset);
>
>
> for (int i=0 ; i<this.compoundURIs.length ; i++) {
> //create data entry
> Individual dataEntry =
> resultDataset.createIndividual(OT.OTClass.DataEntry.getOntClass(resultDataset));
> dataset.addProperty(OT.dataEntry, dataEntry);
> //add compound to data entry
> Individual newcompound =
> resultDataset.createIndividual(this.compoundURIs[i],
> OT.OTClass.Compound.getOntClass(resultDataset));
> dataEntry.addProperty(OT.compound, newcompound);
> //First value
> Individual feature1 =
> resultDataset.createIndividual(MaxtoxApplicationSettings.getServerRootPath()
> + "/feature/" + modelNumber + "_prediction",
> OT.OTClass.Feature.getOntClass(resultDataset));
> Individual featureValue1 =
> resultDataset.createIndividual(OT.OTClass.FeatureValue.getOntClass(resultDataset));
> featureValue1.addProperty(OT.feature, feature1);
> featureValue1.addLiteral(OT.value,
> resultDataset.createTypedLiteral(this.predictions[i],
> XSDDatatype.XSDstring));
>
> //Second value
> Individual feature2 =
> resultDataset.createIndividual(MaxtoxApplicationSettings.getServerRootPath()
> + "/feature/" + modelNumber + "_fingerprint",
> OT.OTClass.Feature.getOntClass(resultDataset));
> Individual featureValue2 =
> resultDataset.createIndividual(OT.OTClass.FeatureValue.getOntClass(resultDataset));
> featureValue2.addProperty(OT.feature, feature2);
> featureValue2.addLiteral(OT.value,
> resultDataset.createTypedLiteral(this.fingerprints[i],
> XSDDatatype.XSDdouble));
>
> //and finally add values to the data entry
> dataEntry.addProperty(OT.values, featureValue1);
> dataEntry.addProperty(OT.values, featureValue2);
> }
> *******************************************************
>
>
>
>
> On 3 September 2010 22:34, Nina Jeliazkova <jeliazkova.nina at gmail.com> wrote:
> > Hi Surajit,
> >
> > I assume resultDataset variable is a Jena Model?
> >
> > Here is one of the RDF files you've tried to upload
> >
> > http://www.ideaconsult.net/downloads/maxtox/ambit2_22761.rdf
> >
> > First observation there is no any ot:Dataset instance. There are multile
> > ot:DataEntry, but no dataset to where they belong. This is definitely a
> > reason the dataset could not be imported - the parser simply doesn't find a
> > dataset to work with and does nothing.
> >
> > Second observation - you have defined value of type double, but the value
> > itself is a string of 0 and 1s (the fingerprint). This is at least not
> > correct and will most probably result in errors when parsing values.
> >
> > <ot:FeatureValue>
> > <ot:value rdf:datatype="http://www.w3.org/2001/XMLSchema#double"
> >
> >>0,0,0,1,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0</ot:value>
> > <ot:feature
> > rdf:resource="http://localhost:8080/MaxtoxTest/feature/1_fingerprint"/>
> > </ot:FeatureValue>
> >
> Corrected that ... I had interchanged the types between prediction and
> fingerprint
>
> > Ah, and (relatively minor) issue - the feature URI
> > http://localhost:8080/MaxtoxTest/feature/1_fingerprint will not be
> > dereferencable, even if you are running feature service at your machine -
> > just because "localhost" means your machine for you, but my machine to me.
> >
>
> Yeah ... that was during testing from my machine ...I have loaded the
> same on the freiburg server which has the correct (absolute) server
> path.
>
> > If you are not running feature service elsewehere it would be better to
> > either have anonymous node for it, or create it beforehand (POST to
> > http://{host}/ambit2/feature) and provide the URI in the dataset RDF. TUM
> > and NTUA developers can share their experience/code as well.
> >
>
> Will try that in a future iteration ... however I have a thought
> regarding this. Remoting is good as long as the information load to be
> maintained is high and the connection is good. I believe localizing
> the smaller stuff is better - as it leads to fewer steps that can fail
> (in the course of a bad connection)
>
> > Third, this code will not work as you might expect:
> >
> > this.uploaded_URI = output.getText(); ,
> >
> > because task RDF will be returned, not a single URI (you have specified
> > request.getClientInfo().getAcceptedMediaTypes().add(new
> > Preference<MediaType>(MediaType.APPLICATION_RDF_XML));
> >
>
> Do I have to drill through the RDF representation to reach the URI ?
> Also if its taking a long time what is your polling strategy ?
>
> Are there any ready pieces of code for polling ?
>
>
>
> > Hope this helps,
> > Nina
> >
>
> Lastly some links I had left out (for reference)
>
> models at : http://opentox2.informatik.uni-freiburg.de:8080/MaxtoxTest/model
> particular (for ex model no 1) :
> http://opentox2.informatik.uni-freiburg.de:8080/MaxtoxTest/model/1
>
> models take a dataset as the input and generate the fingerprints and
> the predictions.
>
> Regards
> Surajit
>
> >
> > On Fri, Sep 3, 2010 at 6:33 PM, surajit ray <mr.surajit.ray at gmail.com>
> > wrote:
> >>
> >> Hi,
> >>
> >> Tried uploading using this piece of code
> >>
> >> ************************************
> >> System.out.println(result_dataset_URI);
> >> Representation representation = new
> >> OutputRepresentation(MediaType.APPLICATION_RDF_XML) {
> >>
> >> @Override
> >> public void write(OutputStream out) throws IOException {
> >> resultDataset.write(out, "RDF/XML-ABBREV");
> >> }
> >> };
> >>
> >> Request request = new Request();
> >> Client client = new Client(Protocol.HTTP);
> >> request.setResourceRef(result_dataset_URI);
> >>
> >> request.setReferrerRef(MaxtoxApplicationSettings.getServerRootPath());
> >> request.setMethod(Method.POST);
> >> request.getClientInfo().getAcceptedMediaTypes().add(new
> >> Preference<MediaType>(MediaType.APPLICATION_RDF_XML));
> >> request.setEntity(representation);
> >>
> >>
> >> Response response = client.handle(request);
> >> // Get the output from the response
> >> Representation output = response.getEntity();
> >> this.uploaded_URI = output.getText();
> >> ************************************
> >>
> >>
> >>
> >>
> >>
> >> I got this as the result
> >>
> >>
> >>
> >>
> >> *************************************
> >> 0.0 Error 1283527568408 File import application/rdf+xml [-1] The
> >> server has fulfilled the request but does not need to return an
> >> entity-body, and might want to return updated metainformation The
> >> server has fulfilled the request but does not need to return an
> >> entity-body, and might want to return updated metainformation 500
> >> *************************************
> >>
> >> Could you clarify ?
> >>
> >> I am using the http://apps.ideaconsult.net:8080/ambit2/dataset for
> >> testing as the other server was down - sorry for that.
> >
> >
> >>
> >> I just uploaded
> >> twice so you can delete 2 vague looking databases ...
> >
> >
> >
> >>
> >> Regards
> >> Surajit
> >>
> >>
> >>
> >> On 3 September 2010 11:46, Nina Jeliazkova <jeliazkova.nina at gmail.com>
> >> wrote:
> >> > Hi Surajit,
> >> >
> >> > On Fri, Sep 3, 2010 at 1:50 AM, surajit ray <mr.surajit.ray at gmail.com>
> >> > wrote:
> >> >>
> >> >> From your previous example I gathered this
> >> >>
> >> >> RDFRepresentation r = new RDFRepresentation();
> >> >> Representation rep = r.createRepresentation(rdfDataset,
> >> >> MediaType.APPLICATION_RDF_XML);
> >> >>
> >> >> The problem is this representation does not exist in the stable
> >> >> version of Restlets !
> >> >
> >> > If you refer to opentox-client examples, then RDFRepresentation is not a
> >> > Restlet class at all, but custom one, using Jena writers for serializing
> >> > RDF.
> >> >
> >> >
> >> > https://ambit.svn.sourceforge.net/svnroot/ambit/branches/opentox/opentox-client/src/main/java/org/opentox/rdf/representation/RDFRepresentation.java
> >> >
> >> > Restlet RDF support is rather minimalistic (graph representation, no
> >> > querying, let alone reasoning), I am not sure if any of the partners
> >> > actually uses it. If you are working with 1.x Restlet , there is indeed
> >> > no
> >> > RDF support there. But Restlet RDF support can be safely ignored if you
> >> > are
> >> > already using Jena (though I would be curious to see performance
> >> > comparison
> >> > between the two).
> >> >
> >> > Using StringRepresentation/ FileRepresentation with media type
> >> > "application/rdf+xml" does the same job. RDFRepresentation class
> >> > discussed
> >> > above actually creates OutputStreamRepresentation.
> >> >
> >> > Upload examples can be found here
> >> >
> >> >
> >> > https://ambit.svn.sourceforge.net/svnroot/ambit/branches/opentox/opentox-client/src/test/java/org/opentox/dataset/test/DatasetTest.java
> >> >
> >> > https://ambit.svn.sourceforge.net/svnroot/ambit/branches/opentox/opentox-client/src/test/java/org/opentox/dataset/test/FeatureTest.java
> >> >
> >> > The examples above are cleaned up versions of ambit code. JUnit tests of
> >> > ambit dataset services can be found here
> >> >
> >> >
> >> > https://ambit.svn.sourceforge.net/svnroot/ambit/trunk/ambit2-all/ambit2-www/src/test/java/ambit2/rest/test/dataset/DatasetsResourceTest.java
> >> >
> >> >
> >> >
> >> >>
> >> >> Also the example uses a RemoteTask to achieve the actual upload.
> >> >> Therefore I could not exactly fathom how to do the upload.
> >> >
> >> > Here is a summary
> >> >
> >> > curl -X POST -d "@filecontent" -H "Content-type:application/rdf+xml"
> >> > http://datasetservice/dataset
> >> >
> >> > This returns Task object (either URI, if -H "Accept:text/uri-list" is
> >> > specified or Task RDF representation if -H "Accept:application/rdf+xml"
> >> > is
> >> > specified). Task RDF representation is defined in
> >> > http://opentox.org/api/1_1/opentox.owl
> >> >
> >> > If the task is still running you'll get result codes 201 or 202 , this
> >> > means
> >> > you need to query again the returned task URI. If code 200 is returned,
> >> > the
> >> > URI returned is where the result should be found. In the examples
> >> > there is
> >> > a RemoteTask class that handles all this.
> >> >
> >> >
> >> > https://ambit.svn.sourceforge.net/svnroot/ambit/branches/opentox/opentox-client/src/main/java/org/opentox/task/RemoteTask.java
> >> >>
> >> >> On 3 September 2010 03:47, surajit ray <mr.surajit.ray at gmail.com>
> >> >> wrote:
> >> >> > Hi,
> >> >> >
> >> >> > I have managed to get a dataset (RDF) made using the results of a
> >> >> > model prediction run. Could you please tell me a dataset service
> >> >> > where
> >> >> > I could upload the same ?
> >> >
> >> > For testing
> >> >
> >> > http://ambit-uni-plovdiv.bg:8080/ambit2/dataset
> >> >
> >> > For your final models/results , especially if you would like them
> >> > appearing
> >> > in ToxPredict.
> >> >
> >> > http://apps.ideaconsult.net:8080/ambit2/dataset
> >> >
> >> >
> >> >>
> >> >> Also it would be very helpful if you could
> >> >> > give an example of an upload request using restlets ...
> >> >
> >> > See above.
> >> >
> >> > Hope this helps,
> >> > Nina
> >> >
> >> >
> >> >> >
> >> >> > Thanx
> >> >> > Surajit
> >> >> >
> >> >
> >> >
> >> >
> >
> >
> >
> > --
> >
> > Dr. Nina Jeliazkova
> > Technical Manager
> >
> >
> > 4 A.Kanchev str.
> > IdeaConsult Ltd.
> >
> >
> > 1000 Sofia, Bulgaria
> > Phone: +359 886 802011
> >
> >
> >
> _______________________________________________
> Development mailing list
> Development at opentox.org
> http://www.opentox.org/mailman/listinfo/development
--
Dipl.-Bioinf. Tobias Girschick
Technische Universität München
Institut für Informatik
Lehrstuhl I12 - Bioinformatik
Bolzmannstr. 3
85748 Garching b. München, Germany
Room: MI 01.09.042
Phone: +49 (89) 289-18002
Email: tobias.girschick at in.tum.de
Web: http://wwwkramer.in.tum.de/girschick
- Previous message: [OTDev] How to upload a dataset ...
- Next message: [OTDev] How to upload a dataset ...
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Development mailing list