The code where swifter is not used will be impacted by swifter

[wanngxiang]

I do not use swifter there. But it raises warnings about swifter. Why?

Environment:
python==3.10
pandas==2.0.3
swifter==1.4.0

[Nimmerfall]

Just having a wild guess (as i don't really know what's going on, but have the same phenomenon)
Is there a call to swifter.set_defaults anywhere in your code?
In my case there is, even though I happen to don't use swifter at all (doesn't seem to work with my use case)

After seeing the intialization code I think what happens is, that an empty DataFrame (or Series) is constructed. I think it's happening to set stuff like partitioning and / or progress bars (see documentation on set_defaults: "NOTE: You must set the defaults before creating the dataframe because this entrypoint is part of the init"

But i can't tell why an empty frame does seem to return True on Frame.index.duplicated().any()

Hope that helps