Issue 14482

wrong georreference statistics?

Reporter: sant
Assignee: omeyn
Type: Feedback
Summary: wrong georreference statistics?
Resolution: Duplicate
Status: Closed
Created: 2013-12-18 10:36:43.402
Updated: 2013-12-19 18:10:02.31
Resolved: 2013-12-18 15:00:56.161
Description: Why does the number of georreferenced records differ between these 2 pages?:

4856 here:

5041 here (which is the real number in our database):


*Reporter*: David G. SanLeĆ³n
*E-mail*: []]]>

Created: 2013-12-18 11:34:27.769
Updated: 2013-12-18 11:34:27.769
There is an issue, but it is not the counts between the "cube" and SOLR, but that we are incorrectly marking things to be geographically suspicious when they are not.

The dataset counts show as 4,856 on:

and when you add the "No Coordinate Issues" flag to the search, the counts in SOLR line up correctly.

The records with issue can be searched too:

This currently returns 185 records, an example of which shows as:

This is clearly marked incorrectly

Created: 2013-12-18 11:37:19.445
Updated: 2013-12-18 11:37:19.445
I suspect that we are being too precise in the geocode-ws, and probably finding a hole not covered between the terrestrial and the EEZ area.

This returns no records:

We need to fix geocode, to return based on a buffer, not a precise lookup

Comment: I suggest closing this issue. The reverse geocode enhancements are addressed in PF-1357
Created: 2013-12-18 14:56:14.239
Updated: 2013-12-18 14:56:14.239

Comment: closed as dupe of PF-1357
Created: 2013-12-18 15:00:56.193
Updated: 2013-12-18 15:00:56.193

Author: sant
Created: 2013-12-19 18:10:02.31
Updated: 2013-12-19 18:10:02.31
Regarding this particular example, now I can't see any records using the 2nd link (search georreferenced occurrences).  It returns "0 occurrences", when it should be returning the same number, or perhaps one more (coordinates added yesterday).

Is this related to corrections in progress, or this is a new different problem?