Grouping by taxon

13902
Reporter: feedback bot
Type: Improvement
Summary: Grouping by taxon
Priority: Minor
Status: Open
Created: 2013-09-13 20:11:47.657
Updated: 2014-01-08 16:46:36.182
        
        
Description: Note: I previously discussed this with Tim and Federico via email, but am putting it here too.

Hello,

From my perspective developing in R, it would be easier if when passing in two keys to a call, eg.

http://uat.gbif.org/occurrence/search?TAXON_KEY=2435099&TAXON_KEY=3119195&limit=100

that this returns 100 results for each key, or up to 100 if less exist, grouped by taxon, instead of returning some mix of both in 100 results total.

I imagine this would be much faster than doing the two separate calls

http://uat.gbif.org/occurrence/search?TAXON_KEY=2435099&limit=100
http://uat.gbif.org/occurrence/search?TAXON_KEY=3119195&limit=100

but I am just assuming two would be slower than 1 call. Perhaps I am wrong?

Federico suggested this could potentially be done, but would require an additional parameter and a different response format.

Thanks! Scott

*Reporter*: Scott Chamberlain
*E-mail*: [mailto:scott@ropensci.org]]]>
    

Author: mdoering@gbif.org
Comment: Is the benefit of creating a new, specific output just performance? Two individual calls would do all that is needed, is that true? In that case I think we should focus on making single calls quicker, rather than creating a very specific response format for a little bit of performance gain. Unless you hit a wall and individual calls are extremely slow and hard to work with
Created: 2013-09-15 12:37:06.908
Updated: 2013-09-15 12:37:06.908