I'm working on #561. With the change in how Pandas handles dates by default (i.e., from a forced ns precision to a more flexible, but now default, us precision), several of our tests have broken, and the output for dates has changed.
That leaves us with two options:
- Understand that the defaults have changed, change tests to adapt to new defaults, while still allowing pandas 2
- Fix everything in a precision
- If we do this, what precision? ns is one option, but us may also make sense. ns can only go out until the 2200s. Some of our climate modeling users may actually find that problematic down the line
For what it's worth, Polars is on us time precision.
I'm working on #561. With the change in how Pandas handles dates by default (i.e., from a forced ns precision to a more flexible, but now default, us precision), several of our tests have broken, and the output for dates has changed.
That leaves us with two options:
For what it's worth, Polars is on us time precision.