I could not understand the principle how G is computed in find_peaks and find_peaks_adaptive , but it works good. Would you explain in more detail?

[xtj49]

In find_peaks_original, G is the mat containing maxima in different scales.
In find_peaks, G needs to be mutiplied a weight.
G = G * np.arange(
N // 2, N // 2 - L, -1
) # normalize to adjust for new edge regions

(1)Why it is a decreasing sequence and how to understand it？
(2)If the sum of G(in axis of 1) means the sum of maxima in different scale, would the weight be [1...L] be right which means the longest length of periodicity?
(3)and would you explain the reason why find_peaks_adaptive works, it is a difficult to understand too