The analysis presented treats the total likelihood as the product of individual likelihoods across frequency bins, verifying that the covariance matrix is diagonal.
In our signal-only analysis, we focus on the signal while discarding the noise, as noise adds to power but does not correlate across frequency bins.
Key to our methodology is ensuring that peak covariances are zero, enabling us to treat them independently for a more straightforward calculation of total likelihood.
Collection
[
|
...
]