This PR aims at improving performance for some of the routes in the API:
Removed the default assumption that all routes need faceting
Removed the default assumption that all routes will have fetch_objects as true
~Changed the way counts are fetched for given bioentity in /bioentity/<type>/<id>. Instead of using countDistinct, we now use cardinality (@kshefchek, thoughts?)~
Changed the way neighbors are fetched in scigraph_util.bioobject method such that only the necessary neighboring nodes are fetched, by traversing a defined relation edge.
Changed the way counts are calculated; now we use a document's association_type
Counts by taxon are also provided, along with total counts
This PR aims at improving performance for some of the routes in the API:
fetch_objects
astrue
/bioentity/<type>/<id>
. Instead of usingcountDistinct
, we now usecardinality
(@kshefchek, thoughts?)~scigraph_util.bioobject
method such that only the necessary neighboring nodes are fetched, by traversing a defined relation edge.association_type
Note: This PR depends on https://github.com/biolink/ontobio/pull/323
New format for counts: