Hi everyone! I know that plotly.py since version 5.16.0 provides some degree of compatibility with dataframe libraries other than pandas via the dataframe interchange protocol. How much compatibility can we expect if a non-pandas library implements the interchange protocol? Is there a list of functionality that would work with the non-pandas dataframe library as well as with pandas, as well as a list of functionality that would work with pandas but now with the non-pandas dataframe library? If we wanted to test for the ability to substitute pandas with a different library, is there a set of integration tests we could run?
I am asking this question in the context of pandas on Snowflake, an extension to Modin that provides a pandas-like dataframe library for using Snowflake.
I appreciate any help! Thank you for reading.
c.f. a related question about scikit-learn: How much compatibility does scikit-learn provide with a dataframe that implements the interchange protocol? · scikit-learn/scikit-learn · Discussion #30077 · GitHub