5.4 Residual diagnostics - identical tables using different methods.

[KonstantinosTsoumas]

Hello again,

Sorry for opening a second discussion so soon.

In the subchapter Portmanteau tests for autocorrelation, the Ljung-Box test output tables for the Naive method and the RandomWalkWithDrift are having the same values. Is this by coincidence or is it a mistake? Given that Random Walk with Drift = previous value + constant drift + noise.

Thanks for your contributions :)

[KonstantinosTsoumas]

In the notebook of this chapter, I only saw it now, the Ljung-Box test results (specifically the tails) are different between the two methods (Naive, RandomWalkWithDrift)

[mergenthaler]

Thanks so much for reporting that discrepancy, @KonstantinosTsoumas. We will take a look.