Here is the new version of the code using
QueryExecution.execConstructTriples:
FileLoader fileLoader = new FileLoader("src/aaCONSTRUCT.tql");
String q = fileLoader.loadAll();
Query query = QueryFactory.create(q) ;
QueryExecution qexec = QueryExecutionFactory.create(query, data.tdb);
Iterator<Triple> ti = qexec.execConstructTriples();
StmtIterator si = ModelUtils.triplesToStatements(ti, data.tdb);
*Now I am trying to add the result of construct in new TDB.*
System.out.println(" ... Add new TDB ...");
inferredData.tdb.add(si);
System.out.println(" ... RDF ...");
inferredData.exportRDF();
inferredData.close();
My problem is that when I do inferredData.tdb.add(si); to save the result
in new TDB (inferredData) the program keeps running for ever and the TDB
size never change 201 MB.
AM I doing something wrong here ?
Thanks
On Thu, Nov 7, 2013 at 6:57 AM, Andy Seaborne <[email protected]> wrote:
> On 07/11/13 02:55, Adeeb Noor wrote:
>
>>
>> On Wed, Nov 6, 2013 at 5:23 AM, Andy Seaborne <[email protected]
>> <mailto:[email protected]>> wrote:
>>
>> On 06/11/13 00:31, Adeeb Noor wrote:
>>
>> Any help with my question please.
>>
>> AdeeB
>>
>>
>> On Mon, Nov 4, 2013 at 1:48 PM, Adeeb Noor
>> <[email protected] <mailto:[email protected]>> wrote:
>>
>> Hi Andy:
>>
>> Thanks for the response.
>>
>> My TDB is on my hard drive with 15GB size wise.
>>
>>
>> How many triples?
>>
>>
>> *The number of triples is: 37397456*
>>
>>
>>
>> And how much of the DB does the SELECT query match?
>>
>> What's SELECT (count(*) AS ?c) ....
>> What's SELECT (count(distinct *) AS ?c) ....
>>
>
> No response?
>
> This is asking what proportion of the DB is being extracted.
>
> On the current information, I'd guess the system starts swapping due to a
> large construct graph but that's just a guess.
>
> Streaming the execConstructTriples to a disk file may help.
>
>
>
>
>>
>> *Here is the SELECT I want to apply to generate my subgraph:*
>>
>
> Long, incomplete query.
>
> Try reordering the FILTERs putting the || ones last. May make no
> difference but without sizing figures, only you can know.
>
> Andy
>
>
>> CONSTRUCT
>>
>> {
>>
>> ddidd:C0004057 ?r ?disease1 .
>>
>> ?disease1 ?r10 ddidd:C0004057 .
>>
>> ?disease1 ?r1 ?omim1 .
>>
>> ?omim1 ?r11 ?disease1 .
>>
>> ?omim1 ?r2 ?w .
>>
>> ?w ?r12 ?omim1 .
>>
>> ?omim1 ?r3 ?bp .
>>
>> ?bp ?r13 ?omim1 .
>>
>> ?omim1 ?r4 ?genotypePhenotype .
>>
>> ?genotypePhenotype ?r14 ?omim1 .
>>
>> ?omim1 ?r5 ?gene.
>>
>> ?gene ?r15 ?omim1 .
>>
>> ?w ?r6 ?gene2.
>>
>> ?gene2 ?r16 ?w .
>>
>> ?omim1 ?r7 ?gene3 .
>>
>> ?gene3 ?r17 ?omim1.
>>
>> ?gene3 ?r8 ?bp2 .
>>
>> ?bp2 ?r18 ?gene3 .
>>
>> ?gene3 ?r9 ?genotypePhenotype2 .
>>
>> ?genotypePhenotype2 ?r19 ?gene3 .
>>
>> ?gene a ?gCLASS.
>>
>> ?gene2 a ?g2CLASS.
>>
>> ?gene3 a ?g3CLASS.
>>
>> ?genotypePhenotype a ?genotypePhenotypeCLASS .
>>
>> ?genotypePhenotype2 a ?genotypePhenotype2CLASS.
>>
>> ?w a ?wCLASS .
>>
>> ?omim1 a ?omimt1 .
>>
>> ?bp a ?bpCLASS .
>>
>> ?bp2 a ?bp2CLASS .
>>
>> ddidd:C0004057 ddids:label ?ldrug1 .
>>
>> ?disease1 ddids:label ?ldisease1 .
>>
>> ?omim1 ddids:label ?lomim1 .
>>
>> ?w ddids:label ?lw .
>>
>> ?bp ddids:label ?lbp .
>>
>> ?genotypePhenotype ddids:label ?lgenotypePhenotype .
>>
>> ?gene ddids:label ?lgene .
>>
>> ?gene2 ddids:label ?lgene2 .
>>
>> ?gene3 ddids:label ?lgene3 .
>>
>> ?bp2 ddids:label ?lbp2 .
>>
>> ?genotypePhenotype2 ddids:label ?lgenotypePhenotype2 .
>>
>> }WHERE{
>>
>> ddidd:C0004057 ?r ?disease1 .
>>
>> ?disease1 ?r10 ddidd:C0004057 .
>>
>> ?disease1 ?r1 ?omim1 .
>>
>> ?omim1 ?r11 ?disease1 .
>>
>> ?omim1 ?r2 ?w .
>>
>> ?w ?r12 ?omim1 .
>>
>> ?omim1 ?r3 ?bp .
>>
>> ?bp ?r13 ?omim1 .
>>
>> ?omim1 ?r4 ?genotypePhenotype .
>>
>> ?genotypePhenotype ?r14 ?omim1 .
>>
>> ?omim1 ?r5 ?gene.
>>
>> ?gene ?r15 ?omim1 .
>>
>> ?w ?r6 ?gene2.
>>
>> ?gene2 ?r16 ?w .
>>
>> ?omim1 ?r7 ?gene3 .
>>
>> ?gene3 ?r17 ?omim1.
>>
>> ?gene3 ?r8 ?bp2 .
>>
>> ?bp2 ?r18 ?gene3 .
>>
>> ?gene3 ?r9 ?genotypePhenotype2 .
>>
>> ?genotypePhenotype2 ?r19 ?gene3 .
>>
>> ?gene a ?gCLASS.
>>
>> ?gene2 a ?g2CLASS.
>>
>> ?gene3 a ?g3CLASS.
>>
>> ?genotypePhenotype a ?genotypePhenotypeCLASS .
>>
>> ?genotypePhenotype2 a ?genotypePhenotype2CLASS.
>>
>> ?w a ?wCLASS .
>>
>> ?omim1 a ?omimt1 .
>>
>> ?bp a ?bpCLASS .
>>
>> ?bp2 a ?bp2CLASS .
>>
>> ddidd:C0004057 ddids:label ?ldrug1 .
>>
>> ?disease1 ddids:label ?ldisease1 .
>>
>> ?omim1 ddids:label ?lomim1 .
>>
>> ?w ddids:label ?lw .
>>
>> ?bp ddids:label ?lbp .
>>
>> ?genotypePhenotype ddids:label ?lgenotypePhenotype .
>>
>> ?gene ddids:label ?lgene .
>>
>> ?gene2 ddids:label ?lgene2 .
>>
>> ?gene3 ddids:label ?lgene3 .
>>
>> ?bp2 ddids:label ?lbp2 .
>>
>> ?genotypePhenotype2 ddids:label ?lgenotypePhenotype2 .
>>
>>
>> FILTER ( ?r = ddids:may_treat || ?r = ddids:may_prevent )
>>
>> FILTER (?omimt1 = ddids:gene || ?omimt1 = ddids:genotypePhenotype )
>>
>> FILTER (?wCLASS = ddids:pathway || ?r2 = ddids:gene_is_element_in_pathway
>> )
>>
>> FILTER (?bpCLASS = ddids:biologicalProcess )
>>
>> FILTER (?bp2CLASS = ddids:biologicalProcess )
>>
>> FILTER (?genotypePhenotypeCLASS = ddids:genotypePhenotype )
>>
>> FILTER (?genotypePhenotype2CLASS = ddids:genotypePhenotype )
>>
>> FILTER (?gCLASS = ddids:gene )
>>
>> FILTER (?g2CLASS = ddids:gene )
>>
>> FILTER (?g3CLASS = ddids:gene )
>>
>> }
>>
>>
>> We still know little about your setup.
>>
>>
>>
>> and my PC is Mac Pro with
>> 2.4 GHZ and 4GB of memory.
>>
>>
>> Java 32 bit or 64 bit?
>>
>>
>> *java version "1.6.0_65"*
>> *Java(TM) SE Runtime Environment (build 1.6.0_65-b14-462-11M4609)*
>> *Java HotSpot(TM) 64-Bit Server VM (build 20.65-b04-462, mixed mode)*
>>
>>
>>
>>
>> I was not able to use QueryExecution.__execConstructTriples
>> as
>>
>> it returnees an iterator and I want to save the subgraph
>> into a new TDB .
>>
>>
>> Why is that a problem? Add them to a TDB database.
>>
>>
>> * I will try it and let you know. *
>>
>>
>>
>> Or even use a SPARQL Update operation.
>>
>>
>> *SPARQL update if I am not wrong will not work in my case as I want to
>>
>> create a subgraph from the whole data and store it in a new TDB. We can
>> use only the SPARQL update if we can add data on the original TDB; AM I
>> right ? *
>>
>> **
>>
>>
>>
>>
>> Here is my code below:
>>
>> FileLoader fileLoader = new
>> FileLoader("src/DDICONSTRUCT.__tql");
>>
>>
>> String q = fileLoader.loadAll();
>>
>> Query query = QueryFactory.create(q) ;
>>
>> QueryExecution qexec =
>> QueryExecutionFactory.create(__query, data.tdb);
>>
>>
>>
>> Model constructModel = qexec.execConstruct();
>>
>>
>> The program has been running for almost a day now, let me
>> know if there is
>> something wrong or if there is an alternative to CONSTRUCT
>> thing.
>>
>>
>>
>> On Sun, Nov 3, 2013 at 12:59 PM, Andy Seaborne
>> <[email protected] <mailto:[email protected]>> wrote:
>>
>> On 03/11/13 07:05, Adeeb Noor wrote:
>>
>> Hi Andy:
>>
>> I did figure it out, however it takes to much time
>> (CONSTRUCT) to finish
>> as
>> my query is complex. Is that something normal ? in
>> fact, it is still
>> running
>>
>>
>> Hard to tell - it depends on many factors such as
>> machine setup, where
>> the data is stored, structure and volume of your data
>>
>> Try
>>
>> QueryExecution.__execConstructTriples
>>
>>
>> Andy
>>
>>
>>
>> AdeeB
>>
>>
>> On Sat, Nov 2, 2013 at 9:56 AM, Adeeb Noor
>> <[email protected]
>> <mailto:[email protected]>>
>>
>> wrote:
>>
>> Hi Andy:
>>
>>
>> Thanks for the quick response. I tried CONSTRUCT
>> and it did work out.
>> But
>> how can I reformat such a query to CONSTRUCT one:
>>
>> SELECT DISTINCT *
>>
>> {
>>
>> ?ddi ddids:has_association ?c .
>>
>> ?ddi ddids:has_association ?c2 .
>>
>> ?c ddids:chemical_or_drug___affects_gene_product
>> ?omim .
>>
>> ?omim ddids:gene_product_encoded_by___gene ?g .
>>
>> ?g ddids:gene_plays_role_in___process ?w .
>>
>> ?g ddids:gene_plays_role_in___process ?bp .
>>
>>
>> ?bp ddids:process_involves_gene ?g2 .
>>
>> ?g2 ddids:gene_plays_role_in___process ?bp2 .
>>
>>
>>
>> where I need each variable ( for example ?w, ?bp
>> , etc) to be a new
>> resources.
>>
>> Thanks
>>
>>
>> On Sat, Nov 2, 2013 at 6:41 AM, Andy Seaborne
>> <[email protected] <mailto:[email protected]>> wrote:
>>
>> You need to use a CONSTRUCT query, not a
>> SELECT one.
>>
>>
>> outputAsRDF encodes the result set (i.e. the
>> table) as RDF - it is not
>> the datamodel of the original data.
>>
>> CONSTRUCT allows you to create one RDF graph
>> from data from another.
>>
>> See also SPARQL Update for doign that from
>> one graph to another in the
>> same database.
>>
>> Andy
>>
>>
>> On 02/11/13 05:35, Adeeb Noor wrote:
>>
>> Hi guys:
>>
>>
>> I would like to save my SPARQL result
>> coming from ResultSet into new
>> rdf.
>> (new rdf resources) cause I want to do
>> more work on this subgraph and
>> it
>> has to be in the original rdf format.
>>
>> I tried outputAsRDF function and it
>> worked however the result I got
>> the
>> following:
>>
>> <rdf:Description rdf:nodeID="A5">
>> <rs:value rdf:resource="
>> https://csel.cs.colorado.edu/~
>> __noor/Drug_Disease_ontology/
>> <https://csel.cs.colorado.edu/
>> ~noor/Drug_Disease_ontology/>
>> DDID.owl#genotypePhenotype
>> "/>
>> <rs:variable>omimt</rs:__
>> variable>
>>
>> </rdf:Description>
>> <rdf:Description rdf:nodeID="A6">
>> <rs:value rdf:resource="
>> https://csel.cs.colorado.edu/~
>> __noor/Drug_Disease_ontology/
>>
>> <https://csel.cs.colorado.edu/
>> ~noor/Drug_Disease_ontology/>
>> DDID.rdf#C0007589
>> "/>
>> <rs:variable>w</rs:variable>
>> </rdf:Description>
>> <rdf:Description rdf:nodeID="A7">
>> <rs:binding rdf:nodeID="A8"/>
>> <rs:binding rdf:nodeID="A9"/>
>> <rs:binding rdf:nodeID="A10"/>
>> <rs:binding rdf:nodeID="A11"/>
>> <rs:binding rdf:nodeID="A12"/>
>> <rs:binding rdf:nodeID="A13"/>
>> <rs:binding rdf:nodeID="A14"/>
>> <rs:binding rdf:nodeID="A15"/>
>> <rs:binding rdf:nodeID="A16"/>
>> <rs:binding rdf:nodeID="A17"/>
>> <rs:binding rdf:nodeID="A18"/>
>> <rs:binding rdf:nodeID="A19"/>
>> <rs:binding rdf:nodeID="A20"/>
>> <rs:binding rdf:nodeID="A21"/>
>> <rs:binding rdf:nodeID="A22"/>
>> <rs:binding rdf:nodeID="A23"/>
>> <rs:binding rdf:nodeID="A24"/>
>> <rs:binding rdf:nodeID="A25"/>
>> <rs:binding rdf:nodeID="A26"/>
>> <rs:binding rdf:nodeID="A27"/>
>> </rdf:Description>
>>
>> how I can remove this nodes things and
>> make it something like:
>>
>> <rdf:Description rdf:about="
>> https://csel.cs.colorado.edu/~
>> __noor/Drug_Disease_ontology/
>>
>> <https://csel.cs.colorado.edu/
>> ~noor/Drug_Disease_ontology/>
>> DDID.rdf#C3229174">
>> <j.0:label>Cytra-K Oral
>> Product</j.0:label>
>> <rdf:type rdf:resource="
>> https://csel.cs.colorado.edu/~
>> __noor/Drug_Disease_ontology/
>>
>> <https://csel.cs.colorado.edu/
>> ~noor/Drug_Disease_ontology/>
>> DDID.owl#chemical
>> "/>
>> </rdf:Description>
>>
>> please help me out
>>
>>
>>
>>
>>
>> --
>> Adeeb Noor
>> Ph.D. Candidate
>> Dept of Computer Science
>> University of Colorado at Boulder
>> Cell: 571-484-3303 <tel:571-484-3303>
>> Email: [email protected]
>> <mailto:[email protected]>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> --
>> Adeeb Noor
>> Ph.D. Candidate
>> Dept of Computer Science
>> University of Colorado at Boulder
>> Cell: 571-484-3303 <tel:571-484-3303>
>> Email: [email protected] <mailto:Adeeb.noor@colorado.
>> edu>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> --
>> Adeeb Noor
>> Ph.D. Candidate
>> Dept of Computer Science
>> University of Colorado at Boulder
>> Cell: 571-484-3303
>> Email: [email protected] <mailto:[email protected]>
>>
>
>
--
Adeeb Noor
Ph.D. Candidate
Dept of Computer Science
University of Colorado at Boulder
Cell: 571-484-3303
Email: [email protected]
