Bug: findPeaks minima branch does not negate data

[iahncajigas]

Bug Description

In SignalObj.m, the findPeaks method's 'minima' branch (line 1630) is identical to the 'maxima' branch (line 1626). The minima case does not negate the data before calling findpeaks, so it returns maxima instead of minima.

Location

SignalObj.m, lines 1624–1631:

if(strcmp(type,'maxima'))
    for i=1:sObj.dimension
        [values{i},indices{i}] = findpeaks(sObj.data(:,i),'MINPEAKDISTANCE',minDistance);
    end
elseif(strcmp(type,'minima'))
    for i=1:sObj.dimension
        [values{i},indices{i}] = findpeaks(sObj.data(:,i),'MINPEAKDISTANCE',minDistance);  % BUG: same as maxima
    end

Expected Behavior

The minima branch should negate the signal before calling findpeaks, then return the original (non-negated) values at those indices:

elseif(strcmp(type,'minima'))
    for i=1:sObj.dimension
        [~,indices{i}] = findpeaks(-sObj.data(:,i),'MINPEAKDISTANCE',minDistance);
        values{i} = sObj.data(indices{i},i);
    end

Found During

Python port (cajigaslab/nSTAT-python) — fixed in the Python implementation.