get_gbif_taxonomy

[Ivypillar]

Hi,

I'm finding this function very useful, however when run at scale, some species names seem to throw an error, and I'm not sure how to resolve them.

For example:

get_gbif_taxonomy("Cediopsylla inaequalis",subspecies=F,higherrank=T)

Error in temp[i] <- taxize::get_gbifid_(temp[[i]]$species[which.max(temp[[i]]$confidence)], :
replacement has length zero

[Ivypillar]

From digging in a little bit I can see that this query gives a genus level match, and that there is a synonym. However I'm still trying to figure out why this breaks something

get_gbifid_("Cediopsylla inaequalis")

Retrieving data for taxon 'Cediopsylla inaequalis'

$Cediopsylla inaequalis
usagekey scientificname rank status matchtype acceptedusagekey
1 1419497 Cediopsylla Jordan, 1925 genus SYNONYM HIGHERRANK 1419561
canonicalname confidence kingdom phylum order family
1 Cediopsylla 94 Animalia Arthropoda Siphonaptera Pulicidae
genus kingdomkey phylumkey classkey orderkey familykey genuskey
1 Spilopsyllus 1 54 216 1366 8853 1419561
class
1 Insecta

[Ivypillar]

The synonym for "Cediopsylla" -> "Spilopsyllus" does seem to have a record on GBIF, however when the internal function get_gbifid_ runs it seems to lose the species epithet, rather than returning "Spilopsyllus inaequalis" as the match.

[fdschneider]

Hi there, I am not actively maintaining the package and it's been a while since it has been developed.
The function get_gbifid_() is provided by the package taxize, and what is returned from the GBIF Taxonomy Server is tricky to handle with simple logical rules, as my wrapper function provides.

As you figured, get_gbifid_(Cediopsylla inaequalis) does not return a 'species' label, and ends with the 'genus' field, which seems highly unusual. This is where the resolved species name should be provided. The missing mapping eventually breaks my function.

The issue seems identical with #35.

The workaround would be to manually replace the species name in the original data with the resolved Synonym. get_gbifid_("Spilopsyllus inaequalis") works and returns a 'status = ACCEPTED'.

[fdschneider]

However: the function get_gbif_taxonomy() should have an exception for this case and return a warning rather than stopping with an Error.

traitdataform/R/get_gbif_taxonomy.R

Line 146 in a1a0b46

temp[i] <- taxize::get_gbifid_(temp[[i]]$species[which.max(temp[[i]]$confidence)], messages = verbose)

Pull requests are welcome.

[Ivypillar]

I have created a pull request with a workaround. It's not an optimal solution as it doesn't rescue the species epithet, but it avoids this specific bug.