AFLPcore
Class TracePeakFinder

java.lang.Object
  |
  +--AFLPcore.TracePeakFinder

public class TracePeakFinder

extends java.lang.Object

This is a simple algorithm that is used to find peaks in a set of trace data. There is a separate file, PeakLocate.java, that was created with this purpose in mind, but that wouldn't have done exactly what I wanted. The algorithm has three steps. 1) Find a point that is high than the points on either side of it and check if it's above the minimum threshold. 2) Check the points thirty samples on either side of it to make sure there's not a nearby point that's even higher. 3) Make sure it meets the minimum width requirement. It's pretty simple, and fairly effective. The peaks on molecular weight standards tend to be very well defined.

Constructor Summary

TracePeakFinder()

Method Summary

static double	find(double[] trace)
static double	find(double[] trace, int offset)
static double	find(double[] trace, int offset, double minPeakValue)
static long	find(double[] trace, int offset, double minPeakValue, double minPeakWidth) This is the one method in this class, really.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

TracePeakFinder

public TracePeakFinder()

Method Detail

find

public static double find(double[] trace)

find

public static double find(double[] trace,
                          int offset)

find

public static double find(double[] trace,
                          int offset,
                          double minPeakValue)

find

public static long find(double[] trace,
                        int offset,
                        double minPeakValue,
                        double minPeakWidth)

This is the one method in this class, really. Quite simple, it searches through the data until it finds a peak, then returns the index of that number. parameters: