occurrence dataset page says '0% with coordinates' although all records have coordinates

[CecSve]

This dataset has coordinates in the records but the page says 0% with coordinates. I ran the archive through the validator and the only issue I see is that the records are flagged presumed negated latitude longitude. Can you take a look @MortenHofft?

[MattBlissett]

I think the percentage is for records with coordinates without issues, i.e. these: https://www.gbif.org/occurrence/search?dataset_key=1451240d-1e30-4aa1-a32f-2231aeb39f80&has_coordinate=true&has_geospatial_issue=false

All the records have a negated longitude.

[MortenHofft]

If you click the file it takes you to the query it performs.
https://www.gbif.org/occurrence/search?dataset_key=1451240d-1e30-4aa1-a32f-2231aeb39f80&has_coordinate=true&has_geospatial_issue=false

So has_geospatial_issue=false
the reason that is the filter is because that is the filter the maps use. That is also the reason the map isn't showing. Because we have a minimum requirement for the coordinates before they show on maps. We could show it as having coordinates, but the map on the dataset page still wouldn't show.
So we do not count what we label as Include records where coordinates are flagged as suspicious

The same goes for years, only if we recognize the year it is counted. For taxa that is more apparent from the text "with match"

[MortenHofft]

But I can see it is confusing. It would also be odd to show that all records had coordinates and then not show the map though. Perhaps a better text or a [?]-popup would be useful

[CecSve]

But I can see it is confusing. It would also be odd to show that all records had coordinates and then not show the map though. Perhaps a better text or a [?]-popup would be useful

Thank you for explaining. I find it a bit odd that the coordinates are showing on the individual occurrence pages (example) and users of the data will also get the records with the interpreted coordinates if they download the dataset archive, but an across-dataset search will exclude the data because they are flagged as suspicious (unless you know how to apply correct filters).

The current way we exclude records with geospatial issues excludes ~4.5 million records for users (and makes sense), but I am not sure the presumed ... flags should be excluded. Perhaps it would make sense to add to the description of the affected fields in the documentation?

[CecSve]

@SophiePamerlon would you be able to assist in updating the longitude coordinates if they are wrong?

@SophiePamerlon never mind - the publisher has corrected the issue.

[CecSve]

But I can see it is confusing. It would also be odd to show that all records had coordinates and then not show the map though. Perhaps a better text or a [?]-popup would be useful

It would be good if the text said that the records have coordinates when we can interpret the data and flag it. But yes, users would probably benefit from a pop-up and perhaps if it is affected by the flag, then the text could have a different colour to highlight something is off?