geoscienceaustralia / auspix_dggs Goto Github PK

The current callablemodules lib uses its own shapefile module. This was introduced as a bit of legacy code. Propose we replace this with pyshp (https://github.com/GeospatialPython/pyshp) and/or Shapely (https://github.com/Toblerity/Shapely), which are well known libraries for the required functionality and well documented/maintained.

When the DGGS resolution is set to a coarse resolution (e.g. 3-5) and auspixdggs.callablemodules.dggs_in_poly_for_geojson_callable is used, depending on the input polygon, the result returned is empty. This is due to the way that the function looks at the centroid of the returned cell and evaluating if it is in the target polygon (which may be off-center).

In this situation, should the result be empty? or should it match the DGGS cell?

On the boundary between the southern S region and the R region (maybe all O P Q R regions) there is a bug in the code. This boundary extents though the top section of Tasmania from west to east. It has been mentioned by users that this only happens at some DGGS resolutions.

Currently the Auspix engine uses WGS84 as the default projection for coordinates. Propose that this can be broadened out to at least GDA94 or the standard datum for Australia.

densify_my_line recently modified to suit only single line
needs to handle both single and multiline

Suggestion to use base16h to represent your grid-cells in a hierarchical geocode...

See illustration here (it runs only on desktop Firefox or Chorme).

It is possible to use positional notation for hierarchical entities. Instead to use base4, base8 or base16 (hexadecimal) you must to use base4h, base8h or base16h.

You can apply it in Hilbert (U) or Morton (Z) space-filling curves.
For example on Hilbert with base16h:

The first (1 digit) and the lest (2 digits) are pure hexadecimal representations. The other are intermediary and half levels... All preserving predixes and its location.