Add check to highlight species on GRIIS missing from be_alientaxa_cube

[SanderDevisscher]

Updating be_alientaxa_cube on Zenodo is a manual operation this creates a possible issue with species not being included in be_alientaxa_cube while they are on the GRIIS checklist. This prompts us to create a workflow to check if there are species missing from
the be_alientaxa_cube. This workflow should be triggered when be_alientaxa_cube is changed on uat.

[soriadelva]

@SanderDevisscher the test has been made but atm it flags 903 taxonkeys as present on the GRIIS checklist and not in the be_alientaxa_cube. Not sure what to do with this 😅. Upon closer inspection, these species are indeed included on the GRIIS Belgium checklist but don't always have occurrence record data that are linked to them. Yet, other species do and are still not included in the cube.

[soriadelva]

@SanderDevisscher the test has been made but atm it flags 903 taxonkeys as present on the GRIIS checklist and not in the be_alientaxa_cube. Not sure what to do with this 😅. Upon closer inspection, these species are indeed included on the GRIIS Belgium checklist but don't always have occurrence record data that are linked to them. Yet, other species do and are still not included in the cube.

By including only species last observed after 1950, 637 remain

[SanderDevisscher]

@soriadelva Is filtering on kingdom an option ? maybe only look at animalia & plantae ?

[soriadelva]

@soriadelva Is filtering on kingdom an option ? maybe only look at animalia & plantae ?

Yes it's an option, that leaves 611

[SanderDevisscher]

@soriadelva still 611 ??

[SanderDevisscher]

can you built in a test to see if the remaining species lack data on gbif (in belgium)? maybe try using rgbif::occ_search(limit=0)$meta$count

[soriadelva]

@soriadelva still 611 ??

Yes

[soriadelva]

can you built in a test to see if the remaining species lack data on gbif (in belgium)? maybe try using rgbif::occ_search(limit=0)$meta$count

Will do 👍

[soriadelva]

rgbif::occ_search(limit=0)$meta$count

When I check if they have occurrences in general and filter those out that haven't, still 601 remain. I'll try for Belgium.

[soriadelva]

rgbif::occ_search(limit=0)$meta$count

When I check if they have occurrences in general and filter those out that haven't, still 601 remain. I'll try for Belgium.

202 left

[SanderDevisscher]

can you provide me the list ?

[soriadelva]

species_not_in_cube.txt

La voilà. Je kunt ook altijd de test al eens runnen op de branch 😉.

[soriadelva]

A quick check shows that there are only a few with a lot of occurrences (>200) in Belgium.

[soriadelva]

The vast majority has only few observations in BE:

[SanderDevisscher]

Maybe exclude preserved specimens (basis of record if i'm not mistaken) Op wo 17 jul. 2024 10:17 schreef soriadelva ***@***.***>:

…

The vast majority has only few observations in BE: Rplot03.png (view on web) <https://github.com/user-attachments/assets/319aa5a0-d383-4d19-a923-f8daa36c9120> — Reply to this email directly, view it on GitHub <#214 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AEAAQODGVNUJV3YAHV23X5LZMYSARAVCNFSM6AAAAABKSY5JROVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMZSG4YTCMJZHA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[soriadelva]

Maybe exclude preserved specimens (basis of record if i'm not mistaken)

After I filtered out preserved specimens and species with only 1 occurrence in Belgium, there are 78 left

[SanderDevisscher]

I can live with that 😄

[soriadelva]

Do we know the rationale about when species are included in the cube, otherwise I can add that to the test.

[SanderDevisscher]

take a look at https://github.com/trias-project/occ-cube-alien

[soriadelva]

I had a look but will need a bit more time to completely grasp what is happening in the aggregate.Rmd file to understand the whole rationale behind the inclusion of the taxa in the cube. I'll pick this up later when there is more time but for now I'll have to put this on hold.

[SanderDevisscher]

I had a look but will need a bit more time to completely grasp what is happening in the aggregate.Rmd file to understand the whole rationale behind the inclusion of the taxa in the cube. I'll pick this up later when there is more time but for now I'll have to put this on hold.

Ok, I suggest we PR the current check & improve it when we have more time to look into it with more detail