[ 
https://issues.apache.org/jira/browse/METRON-1148?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16150737#comment-16150737
 ] 

ASF GitHub Bot commented on METRON-1148:
----------------------------------------

Github user ottobackwards commented on the issue:

    https://github.com/apache/metron/pull/728
  
    Thanks for the contribution!  This looks really nice.
    The question is have is the validation of parameters and the default 
return.  With the FUZZY_SCORE effort, we discussed returning a default 0 score 
when input parameters where wrong, and changed so that errors in parameters 
where known.  This seems to me to be kind of the same thing.  If a non-iterable 
is passed, should it not be an error?


> Add SET and MULTISET data structures to stellar
> -----------------------------------------------
>
>                 Key: METRON-1148
>                 URL: https://issues.apache.org/jira/browse/METRON-1148
>             Project: Metron
>          Issue Type: Improvement
>            Reporter: Casey Stella
>
> With the addition of geohashes, to do analytics like tracking the statistical 
> distribution of the distances of a user's login against the centroid of the 
> user logins across some time, there is a need to be able to store sets (e.g. 
> sets of geohashes) and multisets (sets with multiplicity) in a way that they 
> can be stored by the profiler and merged across time.
> This JIRA should add:
> * SET_INIT
> * SET_ADD
> * SET_REMOVE
> * SET_MERGE
> * MULTISET_INIT
> * MULTISET_ADD
> * MULTISET_REMOVE
> * MULTISET_MERGE
> * MULTISET_TO_SET
> These follow the pattern of the other data structures (that are not stellar 
> language primitives)



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to