Closed mdoering closed 3 months ago
@DianRHR @camiplata I did not check the quality of any of the lists. There are surprisingly many lists from xBio:D in Ohio. Maybe there is a fundamental problem with those?
Full list of all GBIF checklists and their backbone coverage.txt
Thanks! I'll enrich the base coverage data with information about publisher, years, type and other information to have several criteria to sort out the checklist
From the initial list you pointed out we already have several on our reference file, and other still need to be double checked.
Notes on stats generation in Postgres:
create table _md_usages as select dataset_key, count() cnt from name_usage group by dataset_key; create table _md_rels as select dataset_key, count() cnt from nub_rel group by dataset_key; select u.dataset_key, u.cnt, r.cnt as rels, r.cnt*1.0/u.cnt as perc, d.title from _md_usages u left join _md_rels r on r.dataset_key=u.dataset_key left join dataset d on d.key=u.dataset_key where u.cnt > 1000 order by perc;
For this specific issue these were the results:
Consider the following GBIF checklists as xcol sources as they have little overlap with the GBIF backbone or COL: