[okfn-discuss] Guardian article on visualising US public domain data

[Rufus Pollock]

2009/1/22 Jonathan Gray <jonathan.gray at okfn.org>:
> Article about US public domain data, formats, and visualisation.

Thanks for the link.

> # Handy resources for data wonks
> # http://www.guardian.co.uk/help/insideguardian/2009/jan/15/unitedstates-data-journalism-google-spreadsheets
> # Simon Rogers gathers some of the key figures driving the political
> agenda in the United States. Download the data or access the API to
> build your own charts and visualizations.

[snip]

Interesting though I'm not sure 'how handy' a resource this is: a few
hand-picked data sets that haven't been 'normalized' in any way.

For example, while the US public debt item is in fairly standard
tabular form (though with US style dates), the CO2 info has the time
series running the other way (  from left to right) with a load of
additional typical-spreadsheet cruft (blank lines between series etc).

Overall it looks like that they have just copied and pasted from the
originals in which case I'm not sure it is really worth the effort
(why not link to the originals, which in the US are often already in
txt, csv or xls -- though usually 'unnormalized').

A valuable data resource to my mind should either be a collection of
data not available elsewhere, a normalized/packagized version of
existing data, or a (large) collection of existing datasets (also
packaged to some extent).

For example, all the datasets on open economics are either data that
isn't available online to my knowledge (e.g. hand entered from a book)
or has been packaged to provide metadata plus a standardized interface
(e.g. csv/sql + python).

Rufus