Mapping! #27
Description
It would nice to have simple mapping capabilities! Very much a wish list rather than a core requirement.
Users would upload georeferenced datasets formatted like the inputs for Analyse tab but with an additional column or two for GPS coordinates.
The simplest version of this could be very simple:
- base layer using an Open Source map (zooming to smallest area that includes all the sample points)
- points indicating the locations of each sample site
- options to:
- Colour points by presence absence of positive pools
- Colour points by prevalence estimate (PoolPrev)
- Determine size of points by total sample size (number of units or pools)
- Stratify analysis (and visualisation) by a variable or variables and then toggle between strata for visualisation purposes. E.g. if the stratifying variable is Year, you select Year from a list of column names, and then once the stratified prevalence estimates have been run, you can toggle between years in the visualisation
The mapping capabilities could be it's own tab (Home, About, Analyse, Design, Map) or it could be a tab in the output to the 'Analyse' tab (Results, Map, Help). Since the user inputs would be very similar between Analyse and Map, I think the latter makes more sense.
Some additional features that would be cool but more work:
- An option to colour points by an exceedance probability for a given prevalence threshold (relatively easy). This is a way to incorporate uncertainty into a single actionable quantity that can be visualised on a map.
- Coloured polygons rather than points to calculate prevalence in a sampling area based off multiple sampling sites (also easy on the programming side, but users would probably have to supply the polygons and some key to link these to their data which might be limiting step)
- Risk surfaces based on a spatial model. (Harder on the programming side, but we have plans in this direction anyway. This would necessarily be quite slow unfortunately. Online hosting might become an issue in this case)