Heat map for Zimbabwe Idai recovery site

[akahure]

On Zimbabwe recovery site, we would like to include a heat map that will be used to show concentration of sites as different colours based on floods damage and population density.

We would like to see it against population affected, with the latest population data against data from here https://data.humdata.org/dataset/highresolutionpopulationdensitymaps-zwe

The population should correlate to the damaged sites

This is tied to this issue:

[cKellyDesign]

@akahure Please specify which sites layers need heatmap layers.

[cKellyDesign]

This should be a matter of adding a heatmap type layer, which handles data very similarly to symbol layers. In this case we do not have scalable/categorical data on the severity of damage at each site, so these layers should simply show site density.

Given that both heatmap and raster layers are expected to be used at the same time, let's make sure each have partial opacity so we can more easily see the areas of overlap between the two layers.

[cKellyDesign]

As discussed with @ciremusyoka: When we have categorical data for heatmap layers (e.g. damage type is something like "collapsed" or "flooded") should be split into multiple heatmaps layers, one for each category.

The reason for this is because heatmaps are scalar visualizations (in both "density" and "color") which fluidly display a range of data (either by opacity or color) and do not visual categorical data very well. Instead we should should generate multiple heatmap layers with static colors and default density properties (which derive from point density).

Implementation Strategy

Ideally, the functionality to generate the Mapbox stylespec for heatmap layers (regardless of Gisida 1x vs Gisida Views), should:

• [x] 1. Accept parameters for categories and category colors

• [x] 2. Generate styleSpecs for each category

• [x] 3a. [Gisida/-React 1x] Add the array of stylespecs to the layerObj.styleSpec and check for arrays in gisida-react.Map component

• [x] 3b. [Gisida Views] Simply Add an array of heatmap <Layer /> components in the Map render function