Pandas data filtering and parsing speed up

Question

I have the following script I would like to improve in terms of speed,

My data is kind of "huge", with the following shape (1169107, 14); thus the following