Some are using polars for new code, where they might have used pandas before.
I would not call pandas "legacy" though. It it still as ubiquitous and viable as before.
I would not call pandas "legacy" though. It it still as ubiquitous and viable as before.