Verify the approximations for ν given in Equation (416).
In C&E [5] for Section 8.8 we compared two ways of obtaining the EDOFs ν for a WOSA spectral estimator. The first way is via the right-hand approximation from Equation (416), and the second, via Equation (354g), which is based on the WOSA estimator being a special case of a quadratic
spectral estimator (Equation (374a)). Using the same setup as before (a sample size of N = 1024, a Hanning data taper and 50% overlap between blocks) but now considering seven different block sizes, namely, NS = 8, 16, 32, 64, 128, 256 and 512, compute ν using each of the three approximations stated in Equation (416) and also via Equation (354g). How well do these four ways of determining ν match up?
Already registered? Login
Not Account? Sign up
Enter your email address to reset your password
Back to Login? Click here