[OTDev] How to upload a dataset ...
surajit ray mr.surajit.ray at gmail.comFri Sep 3 19:58:22 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 ]
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
>
>
>
- 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