I made a partial correlation plot of the residual of a dataset using python. I got the pacf graph below with some correlation points going to 2 and -1.5. Everyone finds this strange, does someone have any explanation for what is happening?PACF of over 8000 datapoints processed as 68 yearly timesteps.

  • 5
    $\begingroup$ I'm afraid there is a very simple explanation --- it is an error. Sample correlation cannot have a magnitude larger than one. $\endgroup$ – Ben - Reinstate Monica Dec 7 '19 at 0:35
  • $\begingroup$ Thank you @ReinstateMonica. I think during the residual data calculation, I took bias correction of the datapoints as well. I assume that is the most likely point of error. $\endgroup$ – emag_mI Dec 7 '19 at 3:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.