Drop data after stationary and correlation

AI & Machine learning
1

hi everyone,

i am studing the machine learning in trading (quantra book)

at page 37 and 38

after the stationary calculation i find "drop the not stationary data from dataset".

wich function i have to use here?





or do i have to remove the not stationary at the beginnig, after import data for example from yahoo finance?



and,



if after the stationary i have to check the correlation

i have to use the data without not stationary or i have to use the completly data?



is it the same if i remove the not stationary and the correlated data togheter in the function

x.drop(columns…)?!



thank you.

 

2

Hi Irene,



You can remove the non-stationary columns first using the drop() method, then check the correlation for the remaining columns and finally take care of the highly correlated features.



Hope this helps!



Thanks,

Akshay