What happened?
I’ve seen this quite a few times but always during a demo/working session so haven’t had a chance to catch and report it. I think it’s happening only in system generated / cloned encounters that are created when detection runs on an import batch and automatically detects additional animals in an image and creates an encounter.
What we’re seeing is that the country location isn’t getting populated along with the other metadata from the original encounter record / bulk import row:
Hi, Maureen!
We got your bulk import spreadsheet; thank you. However, I couldn’t find the media asset filenames in the encounters in the spreadsheet. E.g., I was looking for CCF_NaChM8_HQElan_2017-Jun-27_23-12-20_20170627-231220-OSRoadE-Playtree318.JPG, but the spreadsheet looks like it has media assets from november, december, and october, but no June. Is it possible that you sent the wrong spreadsheet?
-Mark
Hi @MarkF - sorry about that; I sent the leopard ID spreadsheet instead of the cheetah ID spreadsheet. I’ve just forwarded the latter to the services email.
Hi @colin - thanks for the fix. Does that mean that for all other, pre-existing cloned encounters in the system, we have to fix them manually? And until then, the use of the country as a search filter won’t include results that are currently missing the country? If so that’s thousands of records. This also affects exported datasets.
Can you recommend a way of finding all of these blank country fields in the system currently? I can’t leave the search filter for country blank because that doesn’t return blanks.
Could a script possibly be created to address this? I’m sure ACW isn’t the only wildbook that this bug has affected.
Correct. This clones the encounter country field going forward but does not have any way of correcting already cloned data.
You can’t search only for things that lack a certain field currently, there isn’t an “NOT” clause for the search parameters.
It’s would be technically possible to create a script for this but it’s not a simple one. @tanyastere and I are discussing some possibilities, and evaluating other fields on encounter for whether they should or should not be cloned.