Hi Diego, great article. Thank you !. Just a small query. I am trying to use the OU process for simulating path for EURUSD FX rate. So if I understand correctly: the first portion of the code uptil mean_reverting_procs2.py will be very similar,

However, for the the actual calibration of parameters of the OU process, I will need to plug in historical EUR USD FX rates for the last few years and then run the OLS method (something similar to the mean_reverting_procs_4.py) portion to arrive at the best fit parameters. wherein 'X_t' will be the historical closing rates and 'y' will be the lagged series. Is my understanding correct?. I will mostly use statsmodel for the OLS portion though :)