Feature: Create integration of demographic data sets as Choropleth layers

[kidwellj]

I'm thinking here of the 2011 UK Census, Scottish Index of Multiple Deprivation, and UR 6 fold scale. Once these are integrated then we can start to fill in data in the pane/D3 viz.

[kidwellj]

See good (if slightly more full featured that we would aim for) example here: http://climate.globalforestwatch.org/map/3/15.00/27.00/ALL/dark/biomass_loss?begin=2001-11-10&end=2015-01-01&threshold=30 with code deposited here: or this one, with code here: https://github.com/Vizzuality/gfw-climate

[danhillcode]

progress

[kidwellj]

So here's what needs to be done, I think:

(1) Calculate bounding box for current leaflet map instance displayed onscreen using map.getBounds(); see here: https://stackoverflow.com/questions/22948096/get-the-bounding-box-of-the-visible-leaflet-map#22949388

(1b) Set variables to pass to postgis query based on above bounds (?)

(1c) Set max/min zoom to prevent map from loading too much data at high zoom levels. Basically, > zoom level 4 we should not display any data; between zoom levels 4-8 display a simplified set of polygons

...there are UX considerations here as zoom level will impact the size of data in the query. Options: we could (i) auto-set zoom to minimum if user clicks on an analysis layer or (ii) display light-touch error message to warn them that polygons won't display until zoom level is close enough or (iii) simplify geometries for high zoom levels (see here http://strk.kbt.io/blog/2013/03/08/on-the-fly-simplification-of-topologically-defined-geometries/) or (iv) use alternative data set for high zoom levels.

(2) Query 'demographics' PostGIS database for any objects inside/intersected by this bounding box, something like this I think:

SELECT *
FROM 2011_OAC
WHERE 
  mytable.geom && ST_MakeEnvelope(minLon, minLat, maxLon, maxLat);

see here: https://gis.stackexchange.com/questions/25797/select-bounding-box-using-postgis?rq=1

(3) Display these objects as polygons in a new layer on the leaflet map

An alternative for (1-2) would be to pre-cache all the images using a tileserver like mapnik. I'm pretty sure what these guys here are doing: http://simd.scot/2016/#/simd2016. I think this option may be overly server-heavy and not the most programmatically elegant solution...

[danhillcode]

Just checked DVF library works on its own with the map but struggles with the marker cluster need to look into this more would it be an option to have a separate map with these layers?