[OSGeo Africa] Election maps
Adi Eyal
adi at burgercom.co.za
Sat Mar 23 13:31:04 PDT 2013
Hi Peter
On 23 March 2013 21:42, Peter Newmarch <newmarch at land-surveyors.com> wrote:
> Adi,
>
> I believe this election data has been around for years. The demarcation
> board lists all the ward data, it used to be that one had to ask for it
> specially.
>
Actually I was really surprised at how good the data is. I knew about the
shape files at www.demarcation.gov.za (big up to them for those). The data
on www.elections.gov.za blew me away - every election in the past two
decades, granularity to the level of polling station. Really amazing stuff
- the IEC really gets transparency.
>
> Great work, Fantastic stuff.
>
>
> Some comments - I don’t believe that your 10% criteria is a fair
> reflection of the "safety" of that ward. I believe that a better method
> would be to colour code the winning party, and write its victory total (ie
> 55% of all votes) and its margin of victory in percentage (by ward) - ie if
> the winning party got 55% and the next party got 35%, then it won by 20
> points.
>
Isn't that more or less what I'm doing? The wards are colour coded by ANC,
DA or Other (most parties don't really feature except COPE in the Northern
Cape and IFP in KZN so I don't think it's unfair to bundle them together
for the sake of simplicity). The "at risk" wards are shaded differently to
show that the difference between 1st and 2nd place is less than 10%.
So, for example if the DA got 54% and the ANC got 46% then the difference
is 8 points and I would label that ward as an "at risk" ward.
Actually adding the % votes would be useful but I don't see how to do it
without cluttering the maps. Suggestions?
The one thing that I do think is missing is population. If you look at some
of the municipalities, there are some massive wards and some small ones.
The small ones are usually urban and centred around a town whereas the
large ones are rural with relatively small populations. The maps as
deceptive as you might see a large are labelled with one colour whereas the
a tiny ward might have more impact when counting towards municipal
elections (hope that makes sense)
>
> When a party wins by the 65% upwards range, then it could be considered
> safe.
>
yep - in fact anything above 55% is considered safe as the points
difference can never be less than 10.
Of course their may be better ways, but with a title aimed at 2016 I do
> think it should reflect percentage victories to portray to people the
> possibilities in that data - how safe it is, perhaps with a merger it could
> change, if people of one party dont vote because of a viable alternative -
> who would win etc... - a whole miriad of possibilities.
>
:) now you're getting seriously complex - in theory, a coalition could
destroy a "safe" ward but I don't think you could easily place that
information on a map (unless you have a suggestion)
Thanks for the comments
Adi
>
> Regards
>
> Peter
>
> Peter Newmarch
> Professional Land Surveyor
> 4Y GeoInformatics Pty Ltd
> Tel : +27 31 5642856
> Fax : +27 31 5643074
> Mobile : +27 82 5705859
> eMail : newmarch at land-surveyors.com
>
> On 23/03/2013 21:23, Chris wrote:
>
> I wonder if any politicians are statisticians
> and
> IF any understand attributes and spatial statistics ? ( and forecasting )
>
> Could be a gold mine for someone with the necessary knowledge and skills
>
>
> CM
>
>
>
> On 23 Mar 2013 at 20:38, Adi Eyal wrote:
>
>
> Thanks Tim
>
> Yup - would you believe that the IEC publishes all their data
> in gobsmacking detail? Have a look here: http://www.elections.org.za.
> You'll find the results for all the elections that have taken place over
> the past two decades or so, down to the voting station. I got the
> shapefiles from www.demarcation.gov.za.
>
> Adi
>
> On 23 March 2013 20:35, Tim Sutton <tim at linfiniti.com> <tim at linfiniti.com> wrote:
>
>
> Hi Adi
>
> On Sat, Mar 23, 2013 at 6:03 PM, Adi Eyal <adi at burgercom.co.za> <adi at burgercom.co.za> wrote:
>
> Hi All
>
> I've been fiddling around the election data from the last municipal
> elections and I've put together a graphic showing which wards were
>
> contested
>
> and which are strongholds. I defined contested as there being a less
> than 10% difference in vote counts between the first and second place
> winners.
>
> I'm by no means a cartographer and this is one of my first projects so
> I
>
> was
>
> hoping for comments about how one might improve the maps. I guess the
> one problem is that even though you know that a ward is contested, you
> don't know by how much. Also (and i'm not sure this is important) you
> don't
>
> know
>
> from the map what the individual wards are - only the municipality
> they
>
> are
>
> in. Finally, I don't know where most of these municipalities are
>
> situated so
>
> perhaps so sort of contextual map showing them in relation to the rest
> of the country might be useful. Of course all of these things are
> possible
>
> but
>
> might not be easy to achieve while still maintaining an uncluttered
>
> design.
>
> Any advice would be appreciated. The small version of the image can be
>
> found
>
> here: http://za.okfn.org/2013/03/23/battle-for-local-municipalities/.
>
> The
>
> complete poster is also available as pdf linked to on the site.
>
> Hopefully this sort of question is not frowned upon in this forum (if
> it helps, everything here was produced with open source tools :)).
>
>
> Great project!
>
> Is the source data you used available somewhere publicly? If so can you
> provide a link to it?
>
> Regards
>
> Tim
>
>
>
> Thanks
> Adi
>
> --
> Adi Eyal
> Data Specialist
> phone: +27 78 014 2469
> skype: adieyalcas
> linkedin: http://za.linkedin.com/pub/dir/Adi/Eyal
>
>
> _______________________________________________
> Africa mailing listAfrica at lists.osgeo.orghttp://lists.osgeo.org/mailman/listinfo/africa
>
> --
> Tim Sutton - QGIS Project Steering Committee Member (Release Manager)
> ============================================== Visithttp://linfiniti.com to find out about:
> * QGIS programming services
> * GeoDjango web development
> * FOSS Consulting Services
> Skype: timlinux Irc: timlinux on #qgis at freenode.net
> ==============================================
> _______________________________________________
> Africa mailing listAfrica at lists.osgeo.orghttp://lists.osgeo.org/mailman/listinfo/africa
>
> --
> Adi Eyal
> Data Specialist
> phone: +27 78 014 2469
> skype: adieyalcas
> linkedin: http://za.linkedin.com/pub/dir/Adi/Eyal
>
> _______________________________________________
> Africa mailing listAfrica at lists.osgeo.orghttp://lists.osgeo.org/mailman/listinfo/africa
>
>
>
> _______________________________________________
> Africa mailing list
> Africa at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/africa
>
>
--
Adi Eyal
Data Specialist
phone: +27 78 014 2469
skype: adieyalcas
linkedin: http://za.linkedin.com/pub/dir/Adi/Eyal
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/africa/attachments/20130323/e94f96fb/attachment.html>
More information about the Africa
mailing list