@pacmaninbw Yes I can however, if any little case I can help in the case of accelerating the processing of this/pandas workflow/s on Windows 11, it can be great as my son loves the look and feel of Win 11 and I want to go with it.

Because its a very large dataframe (500 MB) and without multiprocessing the code takes about 35 mins to execute while with Modin it takes about 4 mins. Since I have upgraded to Windows 11, Modin (Ray and Dask) library does not work. I was wondering if multiprocessing can make the job faster. I have a 16 core processor and a 3080 GPU. Multiprocessing can help in this case.

How do I save the resulting dataframe to_csv?

Can I ask for a solution for the same code but using cudf on this platform? How can I ask it?

@greybeard I have updated the code and corrected spelling mistakes. Thank you

Thank you. The code is much cleaner and I get that building functions is a better way to approach this. However, I am still getting the same time as my earlier code and I still get the Boolean Reindexing warning. While I understand, the time cannot be improved however, can I resolve the reindexing warning I get at df = df[~n] and df = df[~o] ?

Thank you for your feedback. I have updated the body of my question. Is there anything else I should be clarifying?