BUG: Timestamp.round floating point error

[jbrockmendel]

• closes #xxxx
• tests added / passed
• Ensure all linting tests pass, see here for how to run them
• whatsnew entry

xref #38964 analogous bug in Timedelta (separate PR upcoming)

ts = pd.Timestamp.max - pd.Timedelta(304)

>>> ts.ceil("s")
Timestamp('1677-09-21 00:12:43.290448384')

[jreback]

ok lgtm. can you add a whatsnew note.

too bad we can't get numpy to actually check for overflow for us. (or is it possible with https://numpy.org/doc/stable/reference/generated/numpy.seterr.html)?

[jbrockmendel]

too bad we can't get numpy to actually check for overflow for us. (or is it possible with https://numpy.org/doc/stable/reference/generated/numpy.seterr.html)?

AFAICT no. numpy/numpy#16352

[jreback]

thanks @jbrockmendel