fabs creates binning bug when tolerance_in_ppm is True

[RogerGinBer]

Hi Pere,

This fabs call in line 75 of the binning makes minMassDistance be always positive, which in turn makes the bin always insert after the current minDistanceIndex (even if the actual distance is negative)

rMSI2/src/peakbinning.cpp

Lines 73 to 78 in 08d7948

	if(tolerance_in_ppm)
	{
	minMassDistance = 1e6*(fabs(minMassDistance)/newMassBin.mass); //Compute distance in ppm
	compTolerance = tolerance;
	newMassBin.binSize = newMassBin.binSize < 0.0 ? (tolerance * newMassBin.mass / 1e6) : newMassBin.binSize; //Overwrite binSize with tolerance when no binSize data is available
	}

rMSI2/src/peakbinning.cpp

Lines 101 to 110 in 08d7948

	if(minMassDistance > 0)
	{
	//Insert the new mass channel after the minDistanceIndex
	targetMassBins.insert(targetMassBins.begin() + minDistanceIndex + 1, newMassBin);
	}
	else
	{
	//Insert the new mass channel before the minDistanceIndex
	targetMassBins.insert(targetMassBins.begin() + minDistanceIndex, newMassBin);
	}

Running some debug tests, this clearly has a huge impact on the resulting data structures and total runtime:

[prafols]

I've just checked it, I agree it was a mistake. Fixed