Author: mdoering@gbif.org
Created: 2012-09-19 15:12:20.169
Updated: 2012-09-19 15:12:20.169
quick look into the number of decendants for these nub taxa shows there are child records inkl species and genera for each of them:
{noformat}
clb-indexing=# select u.id, (rgt-lft-1)/2 as descendants, num_s, num_g, scientific_name from name_usage u join name_string n on name_fk=n.id where u.id in (119,120,121,238,239,4853178,3238258,4836892,4836892,4815623,204) order by 4;
id | descendants | num_s | num_g | scientific_name
---------+-------------+-------+-------+-------------------------
4815623 | 1 | 0 | 1 | Jamoytiidae White, 1946
4853178 | 11 | 9 | 2 | Bothriolepididae
4836892 | 3 | 1 | 2 | Dipnorhynchidae
3238258 | 14 | 8 | 6 | Dipteridae
119 | 98 | 79 | 17 | Myxini
120 | 127 | 71 | 43 | Holocephali
239 | 199 | 60 | 99 | Cephalaspidomorphi
238 | 376 | 106 | 218 | Sarcopterygii
121 | 2039 | 1476 | 462 | Elasmobranchii
204 | 41257 | 32341 | 7648 | Actinopterygii
{noformat}
Author: mdoering@gbif.org
Created: 2012-09-19 15:32:46.995
Updated: 2012-09-19 15:32:55.641
This comes from the fact that we use a limit (100) on facet values. In this example the number of matches in the other fish groups are rather small and less then 100 other higher taxa groups which get returned first. The UI then shows 0 as the count.
A quick fix would be to show "< {minCount}" with minCount being the lowest number of matching records.
We could also increase the facet limit, but not to much more as for each match we need to do a full name usage lookup.
http://staging.gbif.org:8080/checklistbank-search-ws/search/?checklist=nub&highertaxon=121&highertaxon=204&highertaxon=238&facet=highertaxon&facet=checklist
Author: mdoering@gbif.org
Comment: Instead of loading lots of facets immediately we could also load them asynchroneously and also do facet paging. That should speed up searches considerably (removing the need for lots of name usage lookups before rendering). Still getting counts of specific facet values is a different issue
Created: 2012-09-19 15:45:34.606
Updated: 2012-09-19 15:45:34.606
Author: mdoering@gbif.org
Comment: Implementing the min count approach as a start
Created: 2012-09-19 15:46:32.033
Updated: 2012-09-19 15:46:32.033
Author: mdoering@gbif.org
Comment: Showing the < min is not ideal, but at least we dont show anything wrong now. A question mark would also be an option
Created: 2012-09-19 16:03:48.815
Updated: 2012-09-19 16:03:48.815
Author: mdoering@gbif.org
Comment: Decided to close this issue as it has been dealt with. If users find the solution intuitive it should be reopened
Created: 2012-11-21 10:40:16.978
Updated: 2012-11-21 10:40:16.978
