In my last post (Données ouvertes et les émissions de Gaz à effet de serre au Canada) I explained how I convert a CSV data file (from opendata.gc.ca) on greenhouse gases emissions in Canada to build a map. So I produced a tiled map of this project with TileMill deployed on my own server with Wax API. You can try it now !
I tried to show two things 1) the density of facilities producing those emissions 2) the distribution of GHG emissions in Canada. of course I could used a "Heatmap" to show the density of those value. But, I tried to produce this density map using a hexagonal grid.
The idea of presenting the amount of GHGs produced in Canada in a points clustered grid came wit this example (Point Clustering) made by the CartoDB developers. More recently, @elcep also proposed a similar approach in his blog. This concept was perfect to illustrate the distribution of greenhouse gases in a 300km x 300km virtual grid. I builded my layer with this SQL in PostGIS:
SELECT max(install_ges_p.gid) AS gid, sum(install_ges_p.emis_ges) AS gazemis, st_snaptogrid(install_ges_p.the_geom, 0::double precision, 0::double precision, 300000::double precision, 300000::double precision) AS the_geom_gr
FROM install_ges_p
GROUP BY st_snaptogrid(install_ges_p.the_geom, 0::double precision, 0::double precision, 300000::double precision, 300000::double precision);
I loaded the data from my local "PostgreSQL / PostGIS" into my CartoDB account. This tool is cool because it's faster to deploy and doesn't require a tiled map production as everything is dynamic. All we must do is load the table and carto style into your CartoDB account.
@opac_grid:0.1;
@color_gris: #4F1891;
@opac_marker:0.3;
#gridemisgestot[gazemi_nb>0][zoom>=3]{
marker-fill:@color_gris;
marker-opacity:@opac_grid;
marker-line-color:@color_gris;
marker-line-opacity:@opac_marker;
marker-allow-overlap:true;
text-name:"[gazemis]";
text-face-name:"DejaVu Sans Book";
text-allow-overlap:true;
text-halo-radius:1.5;
text-halo-fill:rgba(255,255,255,0.50);
text-placement:point;
[zoom=3]{marker-width:7.5;text-size:4;}
[zoom=4]{marker-width:15;text-size:9;}
[zoom=5]{marker-width:30;text-size:18;}
[zoom=6]{marker-width:60;text-size:36;}
[zoom=7]{marker-width:120;text-size:36;}
}
Then you can
1) use this API
2) or use the small 'embed_map' provided by CartoDB layer by layer.
Not bad. But seems like a complicated way of dealing with a simple matter of displaying data. Comes across as using a hammer to break open an egg.
Not bad. But seems like a complicated way of dealing with a simple matter of displaying data. Comes across as using a hammer to break open an egg.