Libraries such as
promise big speedups to dataframe operations, so we'll want to test their potential to accelerate prx. https://modin.readthedocs.io/en/stable/getting_started/quickstart.html seems especially promising, as it looks like they try to completely emulate the pandas API.