Sofia Heisler No More Sad Pandas 

In this excellent talk, Sofia Heisler explains how to optimize Pandas code. There is a striking difference between a naive implementation and an optimized one, a nearly 500x improvement.

She also described two very useful benchmarking functions, timeit, and line_profiler.

I wonder how much of this would apply to “normal” python development? If nothing else, the two benchmarking functions will certainly come in handy in future optimization.

This entry was posted in data science and tagged . Bookmark the permalink.

Leave a Reply