Hi Markus,
I ran the query against our LOD Cloud Cache server hosting 60+billion LOD
datasets (http://lod.openlinksw.com) including the LGD datasets, and found the
query with Filter A returned 4 results as you indicated returning immediately.
Running the query with Filter B it takes appreciably longer to run and on first
execution gave a timeout error indicating partial results would be returned,
running it again it then 2 results, on the third execution it returned 4
results same as with Filter A . So it would appear the Virtuoso Anytime [1] [2]
query feature is coming into play and returning partial results, thus have you
tried running the query with Filter B multiple times in succession to see if
more results are returned ? see:
http://lod.openlinksw.com/sparql?default-graph-uri=&qtxt=PREFIX+rdfs%3A+%3Chttp%3A%2F%2Fwww.w3.org%2F2000%2F01%2Frdf-schema%23%3E%0D%0APREFIX+ogc%3A+%3Chttp%3A%2F%2Fwww.opengis.net%2Font%2Fgeosparql%23%3E%0D%0APREFIX+geom%3A+%3Chttp%3A%2F%2Fgeovocab.org%2Fgeometry%23%3E%0D%0APREFIX+wgs%3A+%3Chttp%3A%2F%2Fwww.w3.org%2F2003%2F01%2Fgeo%2Fwgs84_pos%23%3E%0D%0APREFIX+lgd%3A++%3Chttp%3A%2F%2Flinkedgeodata.org%2F%3E%0D%0APREFIX+lgdo%3A+%3Chttp%3A%2F%2Flinkedgeodata.org%2Fontology%2F%3E%0D%0APREFIX+lgdm%3A+%3Chttp%3A%2F%2Flinkedgeodata.org%2Fmeta%2F%3E%0D%0APREFIX+lgd-tr%3A+%3Chttp%3A%2F%2Flinkedgeodata.org%2Ftriplify%2F%3E%0D%0APREFIX+lgd-geo%3A+%3Chttp%3A%2F%2Flinkedgeodata.org%2Fgeometry%2F%3E%0D%0APREFIX+dcterms%3A+%3Chttp%3A%2F%2Fpurl.org%2Fdc%2Fterms%2F%3E%0D%0A%0D%0A%0D%0ASELECT+%3Fname%2C+%3Fwkt%2C+bif%3Ast_distance%28%3Fwkt%2C+%3Fcentroid+%29+AS+%3Fdistance%0D%0AFROM+%3Chttp%3A%2F%2Flinkedgeodata.org%3E+%7B%0D%0A+++%3Flgd+a+lgdo%3ACity%3B%0D%0A+++rdfs%3Alabel+%3Fname%3B%0D%0A+++geom%3Ageometry+%5Bogc%3AasWKT+%3Fwkt+%5D.%0D%0A%0D%0A+++BIND+%28+bif%3Ast_point%2810.447683%2C+51.163375%29+AS+%3Fcentroid+%29%0D%0A%0D%0A+++%23+choose+one+of+the+two+lines+below%3A%0D%0A+++%23+FILTER+%28+bif%3Ast_distance%28%3Fwkt%2C+%3Fcentroid%29+%3C%3D+100+%29+%23+use+this+predicate+for+query+A%0D%0A%0D%0A+++FILTER+%28+bif%3Ast_intersects+%28%3Fwkt%2C+%3Fcentroid%2C+100%29+%29+%23+use+this+predicate+for+query+B%0D%0A%0D%0A+++FILTER+%28+langMatches%28lang%28%3Fname%29%2C+%27%27%29+%29%0D%0A+++FILTER+NOT+EXISTS+%7B+%3Flgd+a+lgdm%3AWay.+%7D%0D%0A%7D%0D%0A&format=text%2Fhtml&CXML_redir_for_subjs=121&CXML_redir_for_hrefs=&timeout=30000&debug=on
or short link:
http://bit.ly/1Cdl5rU
[1] http://docs.openlinksw.com/virtuoso/anytimequeries.html
[2]
http://virtuoso.openlinksw.com/dataspace/doc/dav/wiki/Main/VirtTipsAndTricksAnytimeSPARQLQuery
Best Regards
Hugh Williams
Professional Services
OpenLink Software, Inc. // http://www.openlinksw.com/
Weblog -- http://www.openlinksw.com/blogs/
LinkedIn -- http://www.linkedin.com/company/openlink-software/
Twitter -- http://twitter.com/OpenLink
Google+ -- http://plus.google.com/100570109519069333827/
Facebook -- http://www.facebook.com/OpenLinkSoftware
Universal Data Access, Integration, and Management Technology Providers
> On 14 Jan 2015, at 17:38, Markus Ackermann
> <ackerm...@informatik.uni-leipzig.de> wrote:
>
>
> Hi,
>
> testing some geospatial SPARQL queries using a build of the latest commit
> from the developt/7 branch (thus VOS OpenSource 07.10.3211)on the
> LinkedGeoData Dumps from Sep 9 2014 ([1]) I encountered unexpected
> discrepancy
> between the selectiveness of st_intersects versus giving an upper bound to
> the distance between points calculated using st_distance.
>
> I used the following query:
>
> PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
> PREFIX ogc: <http://www.opengis.net/ont/geosparql#>
> PREFIX geom: <http://geovocab.org/geometry#>
> PREFIX wgs: <http://www.w3.org/2003/01/geo/wgs84_pos#>
> PREFIX lgd: <http://linkedgeodata.org/>
> PREFIX lgdo: <http://linkedgeodata.org/ontology/>
> PREFIX lgdm: <http://linkedgeodata.org/meta/>
> PREFIX lgd-tr: <http://linkedgeodata.org/triplify/>
> PREFIX lgd-geo: <http://linkedgeodata.org/geometry/>
> PREFIX dcterms: <http://purl.org/dc/terms/>
>
>
> SELECT ?name, ?wkt, bif:st_distance(?wkt, ?centroid ) AS ?distance
> FROM <http://linkedgeodata.org> {
> ?lgd a lgdo:City;
> rdfs:label ?name;
> geom:geometry [ogc:asWKT ?wkt ].
>
> BIND ( bif:st_point(10.447683, 51.163375) AS ?centroid )
>
> # choose one of the two lines below:
> # FILTER ( bif:st_distance(?wkt, ?centroid) <= 100 ) # use this
> predicate for query A
> # FILTER ( bif:st_intersects (?wkt, ?centroid, 100) ) # use this
> predicate for query B
>
> FILTER ( langMatches(lang(?name), '') )
> FILTER NOT EXISTS { ?lgd a lgdm:Way. }
> }
>
> Using the query A (with filter by st_distance(...) <= 100), I get
> fourcities in my result set.
> Using query B, however, I only get threeout of the fourcities as result.
>
> Please see the Github gist at [2]for details.
>
> All of the four cities that should appear have distances from the
> centroid significantly under
> 100 units, so it shouldn't just be a floating point precision issue or
> something along that lines.
>
> As far as I understand the SQL MM standard and VOS documentation
> st_distance(?a, ?b) <= 100)and st_intersects (?a, ?b, 100)should be
> equivalent predicates
> provided that ?a and ?b are POINT geometries, right?If so, then this
> seems to be a bug.
> Just want to confirm my understanding before creating Github issue on this.
>
>
> Regards,
>
> Markus Ackermann
>
>
>
> [1]http://downloads.linkedgeodata.org/releases/2014-09-09/
>
> [2] https://gist.github.com/neradis/01cb99080baf6e29cde6
>
> ------------------------------------------------------------------------------
> New Year. New Location. New Benefits. New Data Center in Ashburn, VA.
> GigeNET is offering a free month of service with a new server in Ashburn.
> Choose from 2 high performing configs, both with 100TB of bandwidth.
> Higher redundancy.Lower latency.Increased capacity.Completely compliant.
> http://p.sf.net/sfu/gigenet
> _______________________________________________
> Virtuoso-users mailing list
> Virtuoso-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/virtuoso-users
------------------------------------------------------------------------------
New Year. New Location. New Benefits. New Data Center in Ashburn, VA.
GigeNET is offering a free month of service with a new server in Ashburn.
Choose from 2 high performing configs, both with 100TB of bandwidth.
Higher redundancy.Lower latency.Increased capacity.Completely compliant.
http://p.sf.net/sfu/gigenet
_______________________________________________
Virtuoso-users mailing list
Virtuoso-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/virtuoso-users
