[ 
https://issues.apache.org/jira/browse/ARROW-14168?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17452577#comment-17452577
 ] 

Ian Cook edited comment on ARROW-14168 at 12/2/21, 7:42 PM:
------------------------------------------------------------

[~dragosmg] yes, now that {{{}if_else{}}}, {{{}case_when{}}}, and {{coalesce}} 
all natively support factors/dictionaries now, I believe this is now only 
applicable to the aggregate functions. I updated the issue description 
accordingly. In the future there might be other functions that it applies to.


was (Author: icook):
[~dragosmg] yes, now that {{{}if_else{}}}, {{{}case_when{}}}, and {{coalesce}} 
all natively support factors/dictionaries now, I believe this is now only 
applicable to the aggregate functions. In the future there might be other 
functions that it applies to.

> [R] Warn only once about arrow function differences
> ---------------------------------------------------
>
>                 Key: ARROW-14168
>                 URL: https://issues.apache.org/jira/browse/ARROW-14168
>             Project: Apache Arrow
>          Issue Type: Improvement
>          Components: R
>            Reporter: Neal Richardson
>            Assignee: Dragoș Moldovan-Grünfeld
>            Priority: Major
>              Labels: good-first-issue
>             Fix For: 7.0.0
>
>
> When someone calls median or quantile, we warn them that it is approximate. 
> -When someone calls if_else, case_when, or coalesce on a dictionary array, we 
> warn them that it will be converted to string type.- (if_else, case_when, and 
> coalesce now have native support for dictionary arrays.) We probably only 
> need to warn the first time. (This will also solve the issue of uncaught 
> duplicate warnings in our tests, which the testthat 3e upgrade exposes.)



--
This message was sent by Atlassian Jira
(v8.20.1#820001)

Reply via email to