[wdmmg-discuss] Where does the spanish money go

Tue Nov 16 01:13:21 UTC 2010

 Rufus, Anna,

 As a quick proof of concept, I've extracted a subset of the 2010
Spanish budget (1200 rows, top level expense headings) from our full
spreadsheet [1] into a Google Docs one [2], and massaged the data a
bit to match the format of the Israeli sample that was used back in
July [3]. That way, I didn't have to change much from the Israeli
script, just variable names, slice name and translation parameters.
The result is attached, but also available at [4]. (The CKAN package
is missing though, so the load() function won't work.)

 Although I took a look at the wiki [5], I still need to understand a
bit better the structure of the data store (the keys, the
classificationItems...) and how the spanish data fits into it, but
that's possibly best done at the camp with pen and paper. As a quick
summary, just so you understand what's on the spreadsheet:

 - level 1: the budget is split into sections (mostly ministries, but
also things like parliament, supreme court...)
 - level 2: each section is divided into
entities/departments/independent bodies.
 - level 3: each entity gets assigned a number of 'expense programmes'
or goals, like 'primary education' or 'unemployment benefits'. A
programme can be split across multiple entities.
 - level 4: the expenses in a programme are categorised by type:
salaries, investment, utilities... In the spreadsheet I left only the
major headings, which should be enough for now, but in the budget
itself there are 2-3 additional levels of detail.

 So, shall we try to import this into a test slice?

/david

[1]: https://github.com/downloads/dcabo/pge/expenses.xlsx
[2]: https://spreadsheets.google.com/ccc?key=0AowzHU9kHzeudFpFNDJzSi1yR210TTR6dm51dlpZZ0E&hl=en
[3]: http://lists.okfn.org/pipermail/wdmmg-discuss/2010-July/000450.html
[4]: https://github.com/downloads/dcabo/pge/spain.py
[5]: http://wiki.okfn.org/p/Where_Does_My_Money_Go/Store

On Mon, Nov 15, 2010 at 4:56 PM, David Cabo <david.cabo at gmail.com> wrote:
>  Rufus,
>
>  That's great news. I've seen the Israeli spreadsheet, and I can get
> most of the spanish data in that format and adapt the israeli script.
> I'll try to create a Google spreadsheet with a subset of the data
> tonight, and we take it from there.
>
>  Regards,
>
> /david
>
> On Mon, Nov 15, 2010 at 12:56 PM, Rufus Pollock <rufus.pollock at okfn.org> wrote:
>> Hi David,
>>
>> Really good to hear from you. Comments inline below.
>>
>> On 12 November 2010 09:59, David Cabo <david.cabo at gmail.com> wrote:
>> [...]
>>
>>>  I'm very interested in "the money side" of transparency, and I'm a
>>> big fan of wdmmg, so a few weeks ago I started parsing our national
>>> budgets for 2010-2011 and comparing them. Technically the parsing is
>>> not too difficult, as the files are HTML, but the data is presented in
>>> a very fragmented way and without much context. We've got to the point
>>> where we have the data, we're confident it's correct, and we're
>>> thinking about the best way to present it in an easy-to-understand way
>>> (i.e. visualizations). There's not that much to see yet, only a
>>> text-based comparison of the budgets [3] for our own use. (Google
>>> Translate does quite a nice job [4], although it breaks the sortable
>>> tables.)
>>>
>>>  When I looked at the wdmmg front-end code a while ago, the Flash
>>> object was very dependent on the particular structure of the UK
>>> budget. But there was some discussion in this list about having a
>>> generic wdmmg front-end working with a 'pluggable' data store, so I
>>> was wondering if that is still in the roadmap. Since we've got very
>>> limited resources on our side, reusing parts of wdmmg would be
>>> fantastic (and the new version looks great btw).
>>
>> Yes, lots of that work has now been done so the system is much more
>> generic. The simplest thing to do to start with is to write a script
>> to load your data into the where does my money go store (or we can
>> help do that). If you've already got the data in a google spreadsheet
>> this should be pretty easy. For example here's the script for loading
>> some data from israel:
>>
>> <http://bitbucket.org/okfn/wdmmg-ext/src/tip/wdmmgext/load/israel.py>
>>
>> If you'd like more info on this, Anna Powell-Smith, our lead coder on
>> the data side of things (in cc) can probably help.
>>
>>>  I also wanted to say I'll be attending the Open Data Government camp
>>> in London next week, so happy to chat about this or any other matter.
>>> I won't be alone: Jacobo Elosua, who's analysing the budget data and
>>> working on the communication side, will be attending too.
>>
>> It will be great to meet in person!
>>
>> Rufus
>>
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: spain.py
Type: application/octet-stream
Size: 5966 bytes
Desc: not available
URL: <http://lists.okfn.org/pipermail/openspending/attachments/20101116/459a3193/attachment-0001.obj>