-
Notifications
You must be signed in to change notification settings - Fork 1.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Improve numerice_coercion and reuse in both compare op and math op #8385
Conversation
Signed-off-by: jayzhan211 <jayzhan211@gmail.com>
Signed-off-by: jayzhan211 <jayzhan211@gmail.com>
Signed-off-by: jayzhan211 <jayzhan211@gmail.com>
Signed-off-by: jayzhan211 <jayzhan211@gmail.com>
Signed-off-by: jayzhan211 <jayzhan211@gmail.com>
Signed-off-by: jayzhan211 <jayzhan211@gmail.com>
8506795
to
0f6e235
Compare
Signed-off-by: jayzhan211 <jayzhan211@gmail.com>
pub
Thanks @jayzhan211 -- I think this PR needs some careful review to ensure it doesn't introduce subtle bugs / logic errors. I will try and find time but I worry about this PR introducing regressions for no (practical) benefit. Could you please:
Thank you |
If I remember correctly, we currently do not support float16, so it's challenging for us to test these scenarios. |
yes, f16 is not fully support in arrow-rs, so I did not add test |
Which issue does this PR close?
Ref #8302
Rationale for this change
Some type coercion seems incorrect for me, and f16 is not considered.
Something like, but not only
To differentiate the current possible lossy
numeric_coercion
. I think it might be helpful somewhere, but keep ittodo!()
until we need it.What changes are included in this PR?
Are these changes tested?
Since arrow-rs does not support f16 yet, I didn't add test for f16, just ensure the existing test does not fail.
Add math op coercion in slt. some test is not possible to move to slt yet, so keep it there.
Are there any user-facing changes?
additional context
Decimal coercion is different in compare op and math op. Not sure if we can have common coercion for them.