Closed customlogic closed 2 years ago
Dear @customlogic, did you check if the joinBroadPeaks
function? it may suit your needs, it checks if the peaks are closer than a parameter broadWidth
, merge those peaks and optimize the shape parameters of the new peak.
Thanks @jobo322 , I'll look into joinBroadPeaks
After thinking about it more, I guess my problem is that broadenPeaks
creates non-symmetric peaks, but after processing it's impossible to know the offset of "x" within the peak without knowing from
and to
. from
and to
were added to the peak in previous versions, but currently only width
is returned.
@customlogic do you mean the properties left
and right
?
we will back those properties in the structure of Peak1D
Should be fixed in Fixed by https://github.com/mljs/global-spectral-deconvolution/commit/42a54fbd907055d233b9b43c85344823f3551bad
Yes it may lead to non-symmetric peaks.
Thanks, the refactored data structures look really helpful.
Shouldn't broadenPeaks modify the peak's 'x' value in order to prevent overlaps? Currently, it only modifies the width of each peak, but I believe there should be one more step to set the new x value:
The tests for broadenPeak seem to assume the x value will remain unchanged, but the peaks they are expecting still overlap.