[jira] [Comment Edited] (ARROW-13130) [C++][Compute] Add decimal support for arithmetic compute functions

["David Li (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/ARROW-13130?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17421775#comment-17421775 ] 

David Li edited comment on ARROW-13130 at 9/28/21, 8:36 PM:
------------------------------------------------------------

I think we can handle most of these by casting to float64 first (if we want, we could avoid the cast and do the conversion inline instead). abs, round (and ceil, floor, etc), mode, negate, and sign should be handled directly.


was (Author: lidavidm):
I think we can handle most of these by casting to float64 first (if we want, we could avoid the cast and do the conversion inline instead). abs, round (and ceil, floor, etc), mode, negate, sign, and tdigest should be handled directly.

> [C++][Compute] Add decimal support for arithmetic compute functions
> -------------------------------------------------------------------
>
>                 Key: ARROW-13130
>                 URL: https://issues.apache.org/jira/browse/ARROW-13130
>             Project: Apache Arrow
>          Issue Type: New Feature
>          Components: C++
>            Reporter: yibocai#1
>            Assignee: David Li
>            Priority: Major
>              Labels: kernel
>
> The following arithmetic functions do not support decimal:
>  - abs
>  - abs_checked
>  - acos
>  - acos_checked
>  - asin
>  - asin_checked
>  - atan
>  - ceil
>  - cos
>  - cos_checked
>  - floor
>  - is_finite (?)
>  - is_inf (?)
>  - is_nan (?)
>  - ln
>  - ln_checked
>  - log1p
>  - log1p_checked
>  - log2
>  - log2_checked
>  - logb (float/decimal works int/decimal does not)
>  - logb_checked (float/decimal works int/decimal does not)
>  - mode
>  - negate
>  - negate_checked
>  - power (float/decimal works int/decimal does not)
>  - power_checked (float/decimal works int/decimal does not)
>  - quantile
>  - round (ARROW-13975)
>  - round_to_multiple (ARROW-13975)
>  - sign
>  - sin
>  - sin_checked
>  - stddev
>  - tan
>  - tan_checked
>  - tdigest
>  - trunc
>  - variance
> ? - May not be applicable
> The following kernels arithmetic functions do support decimal inputs
>  - add
>  - add_checked
>  - atan2
>  - divide
>  - divide_checked
>  - equal (ARROW-13966)
>  - greater (ARROW-13966)
>  - greater_equal (ARROW-13966)
>  - less (ARROW-13966)
>  - less_equal (ARROW-13966)
>  - mean
>  - min_max
>  - multiply
>  - multiply_checked
>  - product
>  - subtract
>  - subtract_checked
>  - sum
>  - unique



--
This message was sent by Atlassian Jira
(v8.3.4#803005)