It's time to say goodbye to pd.read_csv() and pd.to_csv()

Pandas to CSV input and output operations are serialized, which makes them extremely inefficient and time-consuming. I'm frustrated when I see enough scope for parallelization here, but unfortunately Pandas doesn't offer this yet. Although I'm not a fan of using Pandas to create CSV in the first place (read the article below to find out why), I know there are situations where people have no choice but to use CSV.

Why I stopped dumping DataFrames to CSV and why you should too