更新:到目前为止表现最好的算法是这个。


这个问题探讨了在实时时间序列数据中检测突然峰值的稳健算法。

考虑以下示例数据:

这个数据的例子是Matlab格式的(但这个问题不是关于语言,而是关于算法):

p = [1 1 1.1 1 0.9 1 1 1.1 1 0.9 1 1.1 1 1 0.9 1 1 1.1 1 1 1 1 1.1 0.9 1 1.1 1 1 0.9, ...
     1 1.1 1 1 1.1 1 0.8 0.9 1 1.2 0.9 1 1 1.1 1.2 1 1.5 1 3 2 5 3 2 1 1 1 0.9 1 1, ... 
     3 2.6 4 3 3.2 2 1 1 0.8 4 4 2 2.5 1 1 1];

你可以清楚地看到有三个大峰和一些小峰。这个数据集是问题所涉及的时间序列数据集类的一个特定示例。这类数据集有两个一般特征:

有一种具有一般平均值的基本噪声 有很大的“峰值”或“更高的数据点”明显偏离噪声。

让我们假设以下情况:

峰的宽度不能事先确定 峰的高度明显偏离其他值 算法实时更新(因此每个新数据点都会更新)

对于这种情况,需要构造一个触发信号的边值。但是,边界值不能是静态的,必须通过算法实时确定。


我的问题是:什么是实时计算这些阈值的好算法?有没有针对这种情况的特定算法?最著名的算法是什么?


健壮的算法或有用的见解都受到高度赞赏。(可以用任何语言回答:这是关于算法的)


当前回答

这是一个修改后的Fortran版本的z-score算法。 它是专门针对频率空间中传递函数的峰值(共振)检测进行修改的(每个更改在代码中都有一个小注释)。

如果在输入向量的下界附近存在共振,则第一个修改会向用户发出警告,该共振由高于某个阈值的标准偏差表示(在本例中为10%)。这仅仅意味着信号不够平坦,不足以使检测正确地初始化滤波器。

第二种修改是只将峰值的最大值添加到已找到的峰值中。这是通过将每个发现的峰值与其(滞后)前辈及其(滞后)后继者的大小进行比较来达到的。

第三个变化是尊重共振峰通常在共振频率周围表现出某种形式的对称性。因此,围绕当前数据点对称地计算平均值和std是很自然的(而不仅仅是针对之前的数据点)。这将导致更好的峰值检测行为。

这些修改的效果是,整个信号必须事先被函数知道,这是共振检测的通常情况(类似于Jean-Paul的Matlab示例,其中数据点是动态生成的,这是行不通的)。

function PeakDetect(y,lag,threshold, influence)
    implicit none
    ! Declaring part
    real, dimension(:), intent(in) :: y
    integer, dimension(size(y)) :: PeakDetect
    real, dimension(size(y)) :: filteredY, avgFilter, stdFilter
    integer :: lag, ii
    real :: threshold, influence

    ! Executing part
    PeakDetect = 0
    filteredY = 0.0
    filteredY(1:lag+1) = y(1:lag+1)
    avgFilter = 0.0
    avgFilter(lag+1) = mean(y(1:2*lag+1))
    stdFilter = 0.0
    stdFilter(lag+1) = std(y(1:2*lag+1))

    if (stdFilter(lag+1)/avgFilter(lag+1)>0.1) then ! If the coefficient of variation exceeds 10%, the signal is too uneven at the start, possibly because of a peak.
        write(unit=*,fmt=1001)
1001        format(1X,'Warning: Peak detection might have failed, as there may be a peak at the edge of the frequency range.',/)
    end if
    do ii = lag+2, size(y)
        if (abs(y(ii) - avgFilter(ii-1)) > threshold * stdFilter(ii-1)) then
            ! Find only the largest outstanding value which is only the one greater than its predecessor and its successor
            if (y(ii) > avgFilter(ii-1) .AND. y(ii) > y(ii-1) .AND. y(ii) > y(ii+1)) then
                PeakDetect(ii) = 1
            end if
            filteredY(ii) = influence * y(ii) + (1 - influence) * filteredY(ii-1)
        else
            filteredY(ii) = y(ii)
        end if
        ! Modified with respect to the original code. Mean and standard deviation are calculted symmetrically around the current point
        avgFilter(ii) = mean(filteredY(ii-lag:ii+lag))
        stdFilter(ii) = std(filteredY(ii-lag:ii+lag))
    end do
end function PeakDetect

real function mean(y)
    !> @brief Calculates the mean of vector y
    implicit none
    ! Declaring part
    real, dimension(:), intent(in) :: y
    integer :: N
    ! Executing part
    N = max(1,size(y))
    mean = sum(y)/N
end function mean

real function std(y)
    !> @brief Calculates the standard deviation of vector y
    implicit none
    ! Declaring part
    real, dimension(:), intent(in) :: y
    integer :: N
    ! Executing part
    N = max(1,size(y))
    std = sqrt((N*dot_product(y,y) - sum(y)**2) / (N*(N-1)))
end function std

对于我的应用程序,算法的工作就像一个魅力!

其他回答

下面是在Golang中实现的Smoothed z-score算法(上图)。它假设一个[]int16 (PCM 16bit样本)的切片。你可以在这里找到要点。

/*
Settings (the ones below are examples: choose what is best for your data)
set lag to 5;          # lag 5 for the smoothing functions
set threshold to 3.5;  # 3.5 standard deviations for signal
set influence to 0.5;  # between 0 and 1, where 1 is normal influence, 0.5 is half
*/

// ZScore on 16bit WAV samples
func ZScore(samples []int16, lag int, threshold float64, influence float64) (signals []int16) {
    //lag := 20
    //threshold := 3.5
    //influence := 0.5

    signals = make([]int16, len(samples))
    filteredY := make([]int16, len(samples))
    for i, sample := range samples[0:lag] {
        filteredY[i] = sample
    }
    avgFilter := make([]int16, len(samples))
    stdFilter := make([]int16, len(samples))

    avgFilter[lag] = Average(samples[0:lag])
    stdFilter[lag] = Std(samples[0:lag])

    for i := lag + 1; i < len(samples); i++ {

        f := float64(samples[i])

        if float64(Abs(samples[i]-avgFilter[i-1])) > threshold*float64(stdFilter[i-1]) {
            if samples[i] > avgFilter[i-1] {
                signals[i] = 1
            } else {
                signals[i] = -1
            }
            filteredY[i] = int16(influence*f + (1-influence)*float64(filteredY[i-1]))
            avgFilter[i] = Average(filteredY[(i - lag):i])
            stdFilter[i] = Std(filteredY[(i - lag):i])
        } else {
            signals[i] = 0
            filteredY[i] = samples[i]
            avgFilter[i] = Average(filteredY[(i - lag):i])
            stdFilter[i] = Std(filteredY[(i - lag):i])
        }
    }

    return
}

// Average a chunk of values
func Average(chunk []int16) (avg int16) {
    var sum int64
    for _, sample := range chunk {
        if sample < 0 {
            sample *= -1
        }
        sum += int64(sample)
    }
    return int16(sum / int64(len(chunk)))
}

不需要将极大值与平均值进行比较,还可以将极大值与相邻的最小值进行比较,其中最小值仅定义在噪声阈值之上。 如果局部最大值是>的3倍(或其他置信因子)相邻的最小值,那么这个最大值就是一个峰值。 移动窗口越宽,峰值的确定越准确。 上面使用了以窗口中间为中心的计算, 顺便说一下,而不是在窗口结束时计算(== lag)。

请注意,最大值必须被视为信号之前的增加 之后下降。

下面是平滑z-score算法的Groovy (Java)实现(见上面的答案)。

/**
 * "Smoothed zero-score alogrithm" shamelessly copied from https://stackoverflow.com/a/22640362/6029703
 *  Uses a rolling mean and a rolling deviation (separate) to identify peaks in a vector
 *
 * @param y - The input vector to analyze
 * @param lag - The lag of the moving window (i.e. how big the window is)
 * @param threshold - The z-score at which the algorithm signals (i.e. how many standard deviations away from the moving mean a peak (or signal) is)
 * @param influence - The influence (between 0 and 1) of new signals on the mean and standard deviation (how much a peak (or signal) should affect other values near it)
 * @return - The calculated averages (avgFilter) and deviations (stdFilter), and the signals (signals)
 */

public HashMap<String, List<Object>> thresholdingAlgo(List<Double> y, Long lag, Double threshold, Double influence) {
    //init stats instance
    SummaryStatistics stats = new SummaryStatistics()

    //the results (peaks, 1 or -1) of our algorithm
    List<Integer> signals = new ArrayList<Integer>(Collections.nCopies(y.size(), 0))
    //filter out the signals (peaks) from our original list (using influence arg)
    List<Double> filteredY = new ArrayList<Double>(y)
    //the current average of the rolling window
    List<Double> avgFilter = new ArrayList<Double>(Collections.nCopies(y.size(), 0.0d))
    //the current standard deviation of the rolling window
    List<Double> stdFilter = new ArrayList<Double>(Collections.nCopies(y.size(), 0.0d))
    //init avgFilter and stdFilter
    (0..lag-1).each { stats.addValue(y[it as int]) }
    avgFilter[lag - 1 as int] = stats.getMean()
    stdFilter[lag - 1 as int] = Math.sqrt(stats.getPopulationVariance()) //getStandardDeviation() uses sample variance (not what we want)
    stats.clear()
    //loop input starting at end of rolling window
    (lag..y.size()-1).each { i ->
        //if the distance between the current value and average is enough standard deviations (threshold) away
        if (Math.abs((y[i as int] - avgFilter[i - 1 as int]) as Double) > threshold * stdFilter[i - 1 as int]) {
            //this is a signal (i.e. peak), determine if it is a positive or negative signal
            signals[i as int] = (y[i as int] > avgFilter[i - 1 as int]) ? 1 : -1
            //filter this signal out using influence
            filteredY[i as int] = (influence * y[i as int]) + ((1-influence) * filteredY[i - 1 as int])
        } else {
            //ensure this signal remains a zero
            signals[i as int] = 0
            //ensure this value is not filtered
            filteredY[i as int] = y[i as int]
        }
        //update rolling average and deviation
        (i - lag..i-1).each { stats.addValue(filteredY[it as int] as Double) }
        avgFilter[i as int] = stats.getMean()
        stdFilter[i as int] = Math.sqrt(stats.getPopulationVariance()) //getStandardDeviation() uses sample variance (not what we want)
        stats.clear()
    }

    return [
        signals  : signals,
        avgFilter: avgFilter,
        stdFilter: stdFilter
    ]
}

下面是同一个数据集上的测试,其结果与上面的Python / numpy实现相同。

    // Data
    def y = [1d, 1d, 1.1d, 1d, 0.9d, 1d, 1d, 1.1d, 1d, 0.9d, 1d, 1.1d, 1d, 1d, 0.9d, 1d, 1d, 1.1d, 1d, 1d,
         1d, 1d, 1.1d, 0.9d, 1d, 1.1d, 1d, 1d, 0.9d, 1d, 1.1d, 1d, 1d, 1.1d, 1d, 0.8d, 0.9d, 1d, 1.2d, 0.9d, 1d,
         1d, 1.1d, 1.2d, 1d, 1.5d, 1d, 3d, 2d, 5d, 3d, 2d, 1d, 1d, 1d, 0.9d, 1d,
         1d, 3d, 2.6d, 4d, 3d, 3.2d, 2d, 1d, 1d, 0.8d, 4d, 4d, 2d, 2.5d, 1d, 1d, 1d]

    // Settings
    def lag = 30
    def threshold = 5
    def influence = 0


    def thresholdingResults = thresholdingAlgo((List<Double>) y, (Long) lag, (Double) threshold, (Double) influence)

    println y.size()
    println thresholdingResults.signals.size()
    println thresholdingResults.signals

    thresholdingResults.signals.eachWithIndex { x, idx ->
        if (x) {
            println y[idx]
        }
    }

在Palshikar(2009)中发现了另一个算法:

Palshikar, G.(2009)。时间序列中峰值检测的简单算法。在Proc. 1st Int。高级数据分析,商业分析和智能(卷122)。

论文可以从这里下载。

算法是这样的:

algorithm peak1 // one peak detection algorithms that uses peak function S1 

input T = x1, x2, …, xN, N // input time-series of N points 
input k // window size around the peak 
input h // typically 1 <= h <= 3 
output O // set of peaks detected in T 

begin 
O = empty set // initially empty 

    for (i = 1; i < n; i++) do
        // compute peak function value for each of the N points in T 
        a[i] = S1(k,i,xi,T); 
    end for 

    Compute the mean m' and standard deviation s' of all positive values in array a; 

    for (i = 1; i < n; i++) do // remove local peaks which are “small” in global context 
        if (a[i] > 0 && (a[i] – m') >( h * s')) then O = O + {xi}; 
        end if 
    end for 

    Order peaks in O in terms of increasing index in T 

    // retain only one peak out of any set of peaks within distance k of each other 

    for every adjacent pair of peaks xi and xj in O do 
        if |j – i| <= k then remove the smaller value of {xi, xj} from O 
        end if 
    end for 
end

优势

本文提出了5种不同的峰值检测算法 算法在原始时间序列数据上工作(不需要平滑)

缺点

很难事先确定k和h 峰不能是平的(就像我测试数据中的第三个峰)

例子:

以下是平滑z-score算法的Scala版本(非惯用):

/**
  * Smoothed zero-score alogrithm shamelessly copied from https://stackoverflow.com/a/22640362/6029703
  * Uses a rolling mean and a rolling deviation (separate) to identify peaks in a vector
  *
  * @param y - The input vector to analyze
  * @param lag - The lag of the moving window (i.e. how big the window is)
  * @param threshold - The z-score at which the algorithm signals (i.e. how many standard deviations away from the moving mean a peak (or signal) is)
  * @param influence - The influence (between 0 and 1) of new signals on the mean and standard deviation (how much a peak (or signal) should affect other values near it)
  * @return - The calculated averages (avgFilter) and deviations (stdFilter), and the signals (signals)
  */
private def smoothedZScore(y: Seq[Double], lag: Int, threshold: Double, influence: Double): Seq[Int] = {
  val stats = new SummaryStatistics()

  // the results (peaks, 1 or -1) of our algorithm
  val signals = mutable.ArrayBuffer.fill(y.length)(0)

  // filter out the signals (peaks) from our original list (using influence arg)
  val filteredY = y.to[mutable.ArrayBuffer]

  // the current average of the rolling window
  val avgFilter = mutable.ArrayBuffer.fill(y.length)(0d)

  // the current standard deviation of the rolling window
  val stdFilter = mutable.ArrayBuffer.fill(y.length)(0d)

  // init avgFilter and stdFilter
  y.take(lag).foreach(s => stats.addValue(s))

  avgFilter(lag - 1) = stats.getMean
  stdFilter(lag - 1) = Math.sqrt(stats.getPopulationVariance) // getStandardDeviation() uses sample variance (not what we want)

  // loop input starting at end of rolling window
  y.zipWithIndex.slice(lag, y.length - 1).foreach {
    case (s: Double, i: Int) =>
      // if the distance between the current value and average is enough standard deviations (threshold) away
      if (Math.abs(s - avgFilter(i - 1)) > threshold * stdFilter(i - 1)) {
        // this is a signal (i.e. peak), determine if it is a positive or negative signal
        signals(i) = if (s > avgFilter(i - 1)) 1 else -1
        // filter this signal out using influence
        filteredY(i) = (influence * s) + ((1 - influence) * filteredY(i - 1))
      } else {
        // ensure this signal remains a zero
        signals(i) = 0
        // ensure this value is not filtered
        filteredY(i) = s
      }

      // update rolling average and deviation
      stats.clear()
      filteredY.slice(i - lag, i).foreach(s => stats.addValue(s))
      avgFilter(i) = stats.getMean
      stdFilter(i) = Math.sqrt(stats.getPopulationVariance) // getStandardDeviation() uses sample variance (not what we want)
  }

  println(y.length)
  println(signals.length)
  println(signals)

  signals.zipWithIndex.foreach {
    case(x: Int, idx: Int) =>
      if (x == 1) {
        println(idx + " " + y(idx))
      }
  }

  val data =
    y.zipWithIndex.map { case (s: Double, i: Int) => Map("x" -> i, "y" -> s, "name" -> "y", "row" -> "data") } ++
    avgFilter.zipWithIndex.map { case (s: Double, i: Int) => Map("x" -> i, "y" -> s, "name" -> "avgFilter", "row" -> "data") } ++
    avgFilter.zipWithIndex.map { case (s: Double, i: Int) => Map("x" -> i, "y" -> (s - threshold * stdFilter(i)), "name" -> "lower", "row" -> "data") } ++
    avgFilter.zipWithIndex.map { case (s: Double, i: Int) => Map("x" -> i, "y" -> (s + threshold * stdFilter(i)), "name" -> "upper", "row" -> "data") } ++
    signals.zipWithIndex.map { case (s: Int, i: Int) => Map("x" -> i, "y" -> s, "name" -> "signal", "row" -> "signal") }

  Vegas("Smoothed Z")
    .withData(data)
    .mark(Line)
    .encodeX("x", Quant)
    .encodeY("y", Quant)
    .encodeColor(
      field="name",
      dataType=Nominal
    )
    .encodeRow("row", Ordinal)
    .show

  return signals
}

下面是一个测试,返回与Python和Groovy版本相同的结果:

val y = List(1d, 1d, 1.1d, 1d, 0.9d, 1d, 1d, 1.1d, 1d, 0.9d, 1d, 1.1d, 1d, 1d, 0.9d, 1d, 1d, 1.1d, 1d, 1d,
  1d, 1d, 1.1d, 0.9d, 1d, 1.1d, 1d, 1d, 0.9d, 1d, 1.1d, 1d, 1d, 1.1d, 1d, 0.8d, 0.9d, 1d, 1.2d, 0.9d, 1d,
  1d, 1.1d, 1.2d, 1d, 1.5d, 1d, 3d, 2d, 5d, 3d, 2d, 1d, 1d, 1d, 0.9d, 1d,
  1d, 3d, 2.6d, 4d, 3d, 3.2d, 2d, 1d, 1d, 0.8d, 4d, 4d, 2d, 2.5d, 1d, 1d, 1d)

val lag = 30
val threshold = 5d
val influence = 0d

smoothedZScore(y, lag, threshold, influence)

这里的要点