Update `galah_geolocate` documentation (& functionality?)

[daxkellie]

It might not be clear to users how to use a shapefile with galah_geolocate or how to edit that shape file using the sf package to make it work with galah_geolocate. For example, often shape files are too complex for galah_geolocate, and so it would be worth explaning how to deal with this error with information like:

# If your polygon is too complex, simplify with st_simplify()
library(sf)
sf::sf_use_s2(FALSE)
map <- sf::read_sf("victoria.shp")
map_simplified <- sf::st_simplify(map, dTolerance = 0.05)

[mjwestgate]

galah_geolocate hasn't had much attention, but could probably use some. A few potentially worthwhile improvements might include:

• Using bounding boxes as inputs; though an easier workflow might be to passing inputs to galah_filter via fields such as decimalLatitude and decimallLongitude
• Supporting circular queries by passing lat, long and radius; supported by the 'Get points of interest (POI) within a circle' API, but not called by galah at present
• Allowing users to upload shapefiles to affect later queries; again possible via the API but unclear what workflow would be needed to make this work

[mjwestgate]

Discussion 2022-02-08 led to following proposal:

• galah_polygon() to replace current galah_geolocate functionality, supporting wkt strings. Add complexity limit (11 vertices?)
• galah_radius() to specify a circle as per 'explore your area' tool
• galah_bbox() to extract a bounding box from an sf object or shapefile, and pass that to galah_polygon

One final option (not discussed) is to retain galah_geolocate as a wrapper function to the above arguments. Unclear whether this would be more or less confusing for users.

It would also be good to have a post on ALA labs site (forthcoming) to explain sensible workflows to users

[mjwestgate]

completed with galah_bbox and galah_polygon