search

HughP / olac

Automatically exported from code.google.com/p/olac
0 stars 0 forks source link

Cartogram for OLAC coverage #159

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago 
Generate a cartogram [1,2] showing the number of OLAC items for each
country (the same count listed per country on the region pages, e.g. [3]).
 We probably need to eliminate languages with more than n speakers (n=100
million?).  Language size data will come from Ethnologue.

[1] http://www.worldmapper.org/display_languages.php?selected=583
[2] http://scapetoad.choros.ch/
[3] http://language-archives.org/area/americas

Original issue reported on code.google.com by StevenBird1 on 27 Jan 2010 at 1:25

GoogleCodeExporter commented 9 years ago 
A further restriction is not to count items that lack an OLAC Type, which will
eliminate items that are not data resources (e.g. Ethnologue entries)

Original comment by StevenBird1 on 27 Jan 2010 at 2:04

GoogleCodeExporter commented 9 years ago 
http://arcscripts.esri.com/details.asp?dbid=15638

Original comment by StevenBird1 on 27 Jan 2010 at 11:03

GoogleCodeExporter commented 9 years ago 
Status update:
- The installer downloaded from [4] doesn't install.
- The map data provided by [5] looks good. It's ESRI shapefile.
- ScapeToad downloaded from [2] works, but too slow on my old desktop.
- ScapeToad exports svg and shapefile formats, but without colors.
- uDig [6] can edit shapefiles. It also renders shapefiles as png format.
- uDig can be used for applying a color theme to the ScapeToad output, and then
creating a png image.
- dbfpy [7] is used to add different statistics to the shapefile data set.

Alternatives to ScapeToad
- [8] is developed by the author of the algorithm used by ScapeToad. It's fast 
but
uses bitmap-like files for input output. This can be useful only if we figure 
out how
to create the input format from a shapefile and how to render its output.
- Mapresso [9] can also generate cartograms. Haven't figured out how to upload 
our
own data.

[4] http://arcscripts.esri.com/details.asp?dbid=15638
[5] http://thematicmapping.org/downloads/world_borders.php
[6] http://udig.refractions.net/
[7] http://sourceforge.net/projects/dbfpy/
[8] http://www-personal.umich.edu/~mejn/cart/
[9] http://www.mapresso.com/

Original comment by haepal on 23 Feb 2010 at 10:34

GoogleCodeExporter commented 9 years ago 
Created two cartograms (attached): one including all languages, and the other 
one
without languages with population larger than 10M.

Original comment by haepal on 24 Feb 2010 at 10:37

Attachments:

GoogleCodeExporter commented 9 years ago 
What do the colours mean?

Original comment by StevenBird1 on 10 Feb 2011 at 8:42

GoogleCodeExporter commented 9 years ago 
Colors don't mean anything. Colors were chosen automatically according to a 
scheme that makes it easy to distinguish adjacent regions.

Original comment by haepal on 10 Feb 2011 at 8:47

GoogleCodeExporter commented 9 years ago 
The scheme doesn't appear to work very well!  Note that there are four adjacent 
countries in South America that are the same orange collor.  Are there any 
parameter tweaks available that could ensure that adjacent countries are 
different colors?

Original comment by garyfsim...@gmail.com on 22 Feb 2011 at 12:54