Range values in SDMX

[Ian Dickinson]

If I want to encode an observation which is a range (e.g. "estimated GDP
growth is 1.5% to 2.5%"), how do I do that in SDMX, and more
specifically in SDMX-RDF? I can't see anything in the COG cross-domain
concepts that would allow me to state that a value is a range.

Thanks,
Ian

--
_________________________________________________________
Ian Dickinson i...@epimorphics.com cell: +44-7786-850536
Epimorphics Ltd, Bristol, UK http://www.epimorphics.com
Registered number: 7016688 Tel: +44-1454-410580
Registered: c/o Robson Taylor, Froomsgate House, Bristol

[Spencer Hedger (Nomis)]

Hi,

I've never come across that before, but the only suggestions I could
think of would be to have a "minimum estimated GDP growth" and a
"maximum estimated GDP growth" measure and report it as two
observations? or maybe an "estimated GDP growth" and a "confidence" to
show it could be in that range? Either way from the work I've done
with SDMX I think it would have to be two separate observation
values... of course I could have missed something :)

Spencer.

[Landong Zuo]

Hi Ian,

I had the similar question when I translated the Indices of Multiple
Deprivation data into SDMX-RDF.
I am not an SDMX expert. I am trying to answer the question from OWL/
RDF background as shown in the following example.

Hope it would help. We have a prototype demo available on
http://www.openpsi.org/schools/ based on this data structure.

Please feel free to drop any comment or suggestion.

Regards,

Landong Zuo

Research Fellow,
Electronics and Computer Science,
University of Southampton
mailto:l...@ecs.soton.ac.uk

Here is the example:

@prefix psi: <http://www.openpsi.org/id/> .
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .
@prefix scv: <http://purl.org/NET/scovo#> .
@prefix sdmx: <http://proxy.data.gov.uk/sdmx.org/def/sdmx/> .

<http://www.openpsi.org/def/datatype/Range1.5to2.5> a owl:Class ;
rdfs:label "Percentage range between 1.5 and 2.5" ;
owl:equivalentClass
[ rdf:type rdfs:Datatype ;
owl:onDatatype xsd:float;
owl:withRestrictions (
[ xsd:minIncludsive "0.015"^^xsd:integer ]
[ xsd:maxExcludsive "0.025"^^xsd:integer ]
)
] .

psi:estimatedRange a <http://www.openpsi.org/def/datatype/
Range1.5to2.5> ;
rdfs:label "estimated range";
.

psi:gdpGrowth rdfs:subPropertyOf scv:dimension ;
rdfs:label "GDP Growth" .

psi:estimate rdfs:subPropertyOf rdf:vaule ;
rdfs:label "estimate value" ;
.

psi:GDPGrowth a scv:Dimension ;
rdfs:label "GDP growth" .

psi:gdp_growth000001 a scv:Item ;
a sdmx:Observation;
psi:gdpGrowth psi:GDPGrowth ;
psi:estimate psi:estimatedRange ;
scv:dataset <http://www.openpsi.org/def/imd/gdp/uk/year/2007>;
sdmx:refArea <http://statistics.data.gov.uk/id/country/uk>;
sdmx:timePeriod <http://statistics.data.gov.uk/def/census-year/
2007>;
.

[Ian Dickinson]

Hi Landong,

> I had the similar question when I translated the Indices of Multiple
> Deprivation data into SDMX-RDF.
> I am not an SDMX expert. I am trying to answer the question from OWL/
> RDF background as shown in the following example.
>
> Hope it would help. We have a prototype demo available on
> http://www.openpsi.org/schools/ based on this data structure.
>
> Please feel free to drop any comment or suggestion.

Hmm, that's quite clever, but I wonder how well most RDF processors
could handle it? I'm certain that Jena couldn't do anything except allow
you to navigate the expression at the triple level. You would have to
write the application logic for yourself. I'm not sure if Pellet's
datatype reasoning could infer anything - have you tried?

I guess it could be made more compact if you used the datatype
expression directly as a bNode expression in-line:

psi:gdp_growth000001 a scv:Item, sdmx:Observation;
psi:gdpGrowth [
a [ rdf:type rdfs:Datatype ;

owl:onDatatype xsd:float;
owl:withRestrictions (

[ xsd:minInclusive "0.015"^^xsd:float ]
[ xsd:maxExclusive "0.025"^^xsd:float ]
)
]
];
...

Still, neat idea, thanks.
Ian