Timedelta resolution differs between positional (value, unit) and keyword args

[shiva-sai-824]

Issue #33992 — When constructing a Timedelta, passing the unit as a positional argument vs as a keyword yields different internal resolution and rounding for small fractional inputs.

[shiva-sai-824]

@xcz011 @WillAyd @rhshadrach @mroeschke can you review it and approve?

[jbrockmendel]

Is there supposed to be a non-test change here?

[shiva-sai-824]

Is there supposed to be a non-test change here?

they mentioned that for that test its failing, but when I wrote that test it passed correctly only

[shiva-sai-824]

Is there supposed to be a non-test change here?

if there's no test change and there's change in other main file means, I will change,I just need your confirmation?

[jbrockmendel]

if there's no test change and there's change in other main file means, I will change,I just need your confirmation?

Huh?

[shiva-sai-824]

if there's no test change and there's change in other main file means, I will change,I just need your confirmation?

Huh?

I'm just asking, if there's other file change?I will look into it

[jbrockmendel]

The title and description make it sound like you intend to fix something, not just write a test. But the only code changed so far is a test added. Is the title/description misleading? Or is there a non-test change that needs to be made?

[shiva-sai-824]

The title and description make it sound like you intend to fix something, not just write a test. But the only code changed so far is a test added. Is the title/description misleading? Or is there a non-test change that needs to be made?

That issue is resolved by other, that's why my test is passed
and I added just test case only