Opt out of floor division for float dtype time encoding #9497
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR provides a quick fix for #9488.
Previously for
datetime64[ns]
andtimedelta64[ns]
values we would always use floor division to encode times if we knew that all non-"NaT"
values could be represented with integers with the given units, even if we would ultimately cast the values to floats later based on the specifieddtype
. This causes a problem when encoding"NaT"
values, because the placeholder integer value is converted to a float here, which prevents it from properly being filled with the fill value later in the encoding process.This PR changes things such that if we know that we will convert to floats eventually, we will use floating point division from the start, which preserves
"NaT"
values as floating pointnp.nan
values. I update and test the behavior for bothdatetime64[ns]
values andtimedelta64[ns]
values.cc: @kmuehlbauer
whats-new.rst