Camera Trap DP to Darwin Core

[Victor E. Chocho]

HI, 

Maybe there is a easy way to convert a Agouti Export to Darwin Core for mapping and publishing on GBIF. 

Today I had been hous trying to get Gbif To index a resource and it dindt work, so i had to delete the resourse. 

Wich is the recomend mapping for the terms of the export with the Event, Occurrence and Simple multimedia extensions on the IPT?

[Peter Desmet]

Hi Victor, Xavier and others,

For now, I would recommend the following approach to publishing Camtrap DP datasets to GBIF

1. Export a Camtrap DP from Agouti

2. Load the Camtrap DP with our R package camtraptor, using the function read_camtrap_dp()

3. Check and fix errors

4. Transform the Camtrap DP to Darwin Core with the function write_dwc(). The function documents how the transformation is done: as an Occurrence Core with deployments as parent events and sequences as child events. Media (URLs) are expressed in the Audubon Core Multimedia Extension. Note that it is not possible to have an Event + Occurrence + Media extension, because extensions cannot have extensions in the Darwin Core Archive star schema.

5. Upload the resulting CSV files to an IPT resource

6. Manually add metadata or use the function write_eml() to transform Camtrap DP metadata to EML and upload the EML file.

7. Publish

Here's the entire workflow I have used for this dataset: https://www.gbif.org/dataset/8a5cbaec-2839-4471-9e1d-98df301095dd

I am working with GBIF to make it possible to publish Camtrap DPs directly in the IPT (https://github.com/gbif/ipt/milestone/38). The steps will then be:

1. Export a Camtrap DP from Agouti

2. Upload the datapackage.json and CSV files to the IPT

3. Publish

That workflow isn't available yet, but we hope to release it before the end of this year.

Kind regards,

Peter

--
You received this message because you are subscribed to the Google Groups "agouti-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to agouti-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/agouti-users/67467418-6769-4f0e-a87b-6c7a955262dbn%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Peter Desmet (he/him)

Open data coordinator

Research Institute for Nature and Forest (INBO)

Open science lab for biodiversity

http://orcid.org/0000-0002-8442-8025

[Victor E. Chocho]

Thank you very much, It help me a lot. This is our firts dataset https://www.gbif.org/dataset/2585fcc6-647f-4fb7-a25a-07c515c9d079 

One question, the export csv puts "AGOUTI" as collectionCode, I let that on the published dataset as you can track the use of the platform but I dont know is this was intentional or no.

[Peter Desmet]

Hi Victor,

Nice! Yes, collectionCode is set to "Agouti" for the reason you mention: to be able to track platform across datasets. It is derived from the Camtrap DP "platform" (renamed to sources in the latest version of Camtrap DP), so it is a dynamic value. It is set to "Agouti" by the Agouti export.

Cheers,

Peter

To view this discussion on the web visit https://groups.google.com/d/msgid/agouti-users/deed0862-f633-4c24-b3ff-32b652c34e25n%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[xavier.mestdagh]

Thank you Peter for the details.

All the best,
Xavier