Re: D3 as a possible use case

Hi Ivan,

I'll certainly keep an eye of things from a visualisation angle. I've been
meaning to spend some time going over the existing CSV WG docs, but haven't
had time just yet.

D3 is incredibly good at mapping data to visuals, but people often program
D3 in an imperative way (because of the temptation to program imperatively
in javascript) so one of the challenges will be making a declarative
visualisation mapping (e.g. using RDF) that people feel comfortable with.
I'm using the Grammar of Graphics to provide a declarative layer on top of
D3 which gives some clear mappings about how data will be transformed to
visuals. I will have something to show on this front in a few weeks.

The Vega project [1] also puts a declarative mapping on top of D3 so might
be worth a look.

Thanks,

Kev

[1] https://github.com/trifacta/vega/wiki

On 12 September 2014 09:52, Ivan Herman <ivan@w3.org> wrote:

>
> On 11 Sep 2014, at 18:39 , Kev Kirkland <kev@dataunity.org> wrote:
>
> > Hi guys,
> >
> > D3 is very flexible so you can use things like raw arrays of integers
> for visualisations, but the most common way of formatting data is to have
> an array of JSON/JavaScript objects, e.g:
> > [
> >   {
> >     "Field1": 50,
> >     "Field2": 55
> >   },
> >   {
> >     "Field1": 99,
> >     "Field2": 90
> >   }
> > ]
> >
> > When you use the D3 CSV parsing, it will generate a list of Javascript
> objects like the above by default.
> >
> > D3 has data accessor functions which let you specify how to pull figures
> out of your data, e.g for the data above you might use something like the
> following to access the figures in Field1:
> >
> > function(d) { return d.Field1; }
> >
> > The parameter 'd' represents a row of data, so you can use any
> javascript expressions to access and prepare the data.
> >
> > I've been meaning to drop an email to the list. This is one of the areas
> I'm interested in.
>
> Great! So keeping an eye on what we discuss and do with the questions like
> "Does this help/hinder a D3 application" would be great!
>
> Thanks
>
> Ivan
>
>
> > I'm currently working on a project that puts a Grammar of Graphics layer
> on top of D3. I'm modelling this layer in RDF so it connects to my data
> pipeline processing vocabulary (which in turn connects to DCAT).
> >
> > I've been reusing elements of RDF Data Cube for my metadata
> specifications at the moment, but naturally I'm very interested in the work
> coming out of CSV WG.
> >
> > Thanks,
> >
> > Kev
> >
> > On 11 September 2014 17:03, Ivan Herman <ivan@w3.org> wrote:
> > Jeremy, guys,
> >
> > I had a chat today with a colleague who uses D3[1]. What he told me
> that, in fact, D3 can also directly input CSV data. We were wondering (and
> neither of us knew) whether D3 has or not some extra requirements on the
> type of data they use, the way they convert the CSV data in whatever format
> they use internally (I think it is JSON). I realize this is not the usual
> use case, but we were wondering whether this would be a very special but
> nevertheless useful use case.
> >
> > The two persons that seem to be good resources here are:
> >
> > Scott Murray <scott.murray@alignedleft.com>
> > Mike Bostock <mbostock@gmail.com>
> >
> > Cheers
> >
> > Ivan
> >
> >
> > [1] http://d3js.org
> >
> > ----
> > Ivan Herman, W3C
> > Digital Publishing Activity Lead
> > Home: http://www.w3.org/People/Ivan/
> > mobile: +31-641044153
> > GPG: 0x343F1A3D
> > WebID: http://www.ivan-herman.net/foaf#me
> >
> >
> >
> >
> >
> >
> >
> >
> > --
> > www.dataunity.org
> > twitter: @data_unity
>
>
> ----
> Ivan Herman, W3C
> Digital Publishing Activity Lead
> Home: http://www.w3.org/People/Ivan/
> mobile: +31-641044153
> GPG: 0x343F1A3D
> WebID: http://www.ivan-herman.net/foaf#me
>
>
>
>
>
>


-- 
www.dataunity.org
twitter: @data_unity