In which Wildbook did the issue occur? Manta Matcher
What operating system were you using? MacOS Mojave 10.14.6
What web browser were you using? Chrome Version 86.0.4240.80
What is your role on the site? username argeswaraj; manager, researcher, reviewer, local, image processor, approve, identifier, adoption, Indonesia (all locations)
What happened? I’ve been using the Bulk upload template for a couple of times now on IoT and Manta Matcher. It’s been going well, however there were a few things:
The ID number when uploaded to MM became a new individual instead of linking to the existing individual. For instance, I put in INNLP0512A on the bulk sheet, and when it is uploaded, instead of being linked to the existing INNLP0512A individual page, it comes out as if it’s a new individual and only has one encounter on the page.
The Encounter.distinguishingScar column on the bulk sheet template isn’t detected on MM, so all inputs in the column aren’t visible on the encounter page
I asked about multiple keywords in one picture through @colin on an email, and he said to separate each keyword with a (_). I tried separating it on the bulk sheet, but instead of being individual keywords, it ended up as one whole keyword.
Colin told me this was an issue on all wild book, where media asset 2 and 3 and their keywords cannot be detected
What did you expect to happen?
the IDs are linked to the existing individual page instead of creating a new page
scarring comments are visible on the encounter page
be able to have several keywords in a picture.
ability to upload multiple pictures in one encounter using the bulk upload
What are some steps we could take to reproduce the issue?
Help me outtt with the bulk upload issues
We are starting to take a look at this and will keep you updated.
I can answer #2 right away though: Distinguishing scar isn’t a text field. It was designed just as a positive/negative and will accept ‘1’ for distingushing scar is present, and 0 for not present.
We’ll be working with the excel sheet you provided and be in touch about points 1 and 3.
I am having similar issues when attempting to bulk upload to MantaMatcher.
Do you know what the correct field name would be for the “Noticeable scarring” field on the MantaMatcher encounter reports?
For me, it is Encounter.mediaAsset0 and Encounter.mediaAsset1 columns that aren’t being recognised. I have emailed the spreadsheet to services@wildme.org in case you need it.
Hey all,
regarding the Noticeable scarring field issue (#2): @StephVenables, the field for Noticeable scarring is Encounter.distinguishingScar on all platforms. However, there’s a bug in the bulk import that we need to address. @argeswaraj The field changed from a numeric entry (0 or 1 to indicate no or some scarring) to the description field you see in the UI. However, the bulk import is still trying to process the numeric. We’re tracking this under WB-1107 and are working to get it resolved quickly.
I’m still working through the other points in your post, but wanted to let you know that that one is definitely an issue on our end!
@argeswaraj,
I believe I’ve figured out the keyword issue (#3) you were experiencing. We have two different methods of supporting keywords, and only one of them allows entering multiple keywords in a single cell. Looking at your spreadsheet, you were using the one that supports single keywords.
I recommend updating your spreadsheet to use Encounter.mediaAssetX.keywords. With this, you enter all keywords associated with a picture in a single cell with the keywords separated by _.
I just uploaded this:
And got this in the UI:
So that should be all you need to fix this.
You can cleanly import your data for those three columns be removing the leading space (" “) in the title, for example " Encounter.mediaAsset0.keywords” needs to be “Encounter.mediaAsset0.keywords”
I also recommend changing your images the JPG rather than PNG to ensure we can help match them. We may support PNG for matching (I don’t honestly know), but JPG is the standard.
For me, it is Encounter.mediaAsset0 and Encounter.mediaAsset1 columns that aren’t being recognised. I have emailed the spreadsheet to services@wildme.org in case you need it.
Hi @StephVenables, we see “Encounter.mediaAsset0” not import when part of the MantaMatcher pipeline gets stale. Restarting the server fixes it. It’s an issue we’re still tracking down.
I have restarted MantaMatcher, and I now see Encounter.mediaAsset0 importable with your spreadsheet. If you see this again, just ask for a restart here. It does occasionally creep up.
The ID number when uploaded to MM became a new individual instead of linking to the existing individual. For instance, I put in INNLP0512A on the bulk sheet, and when it is uploaded, instead of being linked to the existing INNLP0512A individual page, it comes out as if it’s a new individual and only has one encounter on the page.
Unfortunately, issue number 1 (ID written on the bulk upload sheet turns into a new individual when uploaded and not linked to an existing manta) is still an issue.
Bulk upload has been making things easier, but with this issue, it makes assigning the encounter to a manta twice the work as we have to edit the encounter page to add the actual ID.
