Idea: Add a simple preview map in the IPT management interface

[timrobertson100]

In light of this this discussion I wonder if it would be useful for the IPT to generate a simple image that shows the verbatim coordinates contained in the dataset, after the DwC-A is produced?

It doesn't need to be fancy, but a simple visual that shows where the Lat/Lng are located. This is not a single dataset and just to show an example, but if the IPT produced something like this, a publisher might be able to spot issues very easily:

(one could also generate a histogram of the year counts, a histogram of the month counts, a table of counts of kingdom, phylum etc too)

[peterdesmet]

As long as it doesn't require us to install Geoserver (cf. IPT v1). 😅

[MattBlissett]

When I've generated maps like these for analytics, data quality discussions and so on I've used this script with small modifications: https://github.com/gbif/analytics/blob/master/R/snapshot_density_map.py (I'm sure there are nicer Python ways to do this, and a straightforward Java way for the IPT.)

A cell size of 1° is a good balance; much smaller and it becomes difficult to see outliers, larger and the map becomes a large blob.

[timrobertson100]

Accumulating at 1 deg and producing a GeoJSON file might be a simple solution. It can then just overlaying using our tile service for the base layer on any slippy map client.

[timrobertson100]

I knocked up a super quick example that reads the DwC-A, and accumulates some metrics. An example GeoJSON is here

It looks like reading a million rows from the DwC-A is very quick and might be an easy way to implement some metrics and views.