This PR fixes the issue that during the computation of a sample correlation, the longer trace wasn't demeaned with the corresponding local mean. Instead, it was assumed to be demeaned overall, which is incorrect.
I'm not sure if the provided fix can be further optimized.
@luca-s, would you be so kind and check if the fix works properly? Thanks.
This PR fixes the issue that during the computation of a sample correlation, the longer trace wasn't demeaned with the corresponding local mean. Instead, it was assumed to be demeaned overall, which is incorrect.
I'm not sure if the provided fix can be further optimized.
@luca-s, would you be so kind and check if the fix works properly? Thanks.