如何在NumPy中高效地找到平滑多维数组的局部最小值？

25 浏览
0 Comments

如何在NumPy中高效地找到平滑多维数组的局部最小值？

假设我有一个NumPy数组，其中包含对连续可微函数的评估，我想找出局部极小值。没有噪音，因此，其值小于所有邻居的值的每个点都符合我的局部最小值标准。

我有以下列表推导式，该式适用于二维数组，不考虑边界上的潜在最小值:

import numpy as N
def local_minima(array2d):
    local_minima = [ index 
                     for index in N.ndindex(array2d.shape)
                     if index[0] > 0
                     if index[1] > 0
                     if index[0] < array2d.shape[0] - 1
                     if index[1] < array2d.shape[1] - 1
                     if array2d[index] < array2d[index[0] - 1, index[1] - 1]
                     if array2d[index] < array2d[index[0] - 1, index[1]]
                     if array2d[index] < array2d[index[0] - 1, index[1] + 1]
                     if array2d[index] < array2d[index[0], index[1] - 1]
                     if array2d[index] < array2d[index[0], index[1] + 1]
                     if array2d[index] < array2d[index[0] + 1, index[1] - 1]
                     if array2d[index] < array2d[index[0] + 1, index[1]]
                     if array2d[index] < array2d[index[0] + 1, index[1] + 1]
                   ]
    return local_minima

然而，这很慢。我还想让它适用于任意维度的数组。例如，有没有一种简单的方法可以获取数组中任意维度点的所有邻居？或者我完全错误地处理了这个问题？我应该使用numpy.gradient()替代吗？

admin 更改状态以发布 2023年5月24日
0

2 答案

0 Comments

尝试使用以下方法获取2D数组：

import numpy as N
def local_minima(array2d):
    return ((array2d <= N.roll(array2d,  1, 0)) &
            (array2d <= N.roll(array2d, -1, 0)) &
            (array2d <= N.roll(array2d,  1, 1)) &
            (array2d <= N.roll(array2d, -1, 1)))


该方法将返回一个类似于数组的二维数组，其中包含真/假值，其中包含局部极小值（四个邻居）。

0
0 Comments

通过对一个任意维度的数组使用Ivan的detect_peaks函数，可以找到局部最小值的位置，需要进行轻微修改:

import numpy as np
import scipy.ndimage.filters as filters
import scipy.ndimage.morphology as morphology
def detect_local_minima(arr):
    # https://stackoverflow.com/questions/3684484/peak-detection-in-a-2d-array/3689710#3689710
    """
    Takes an array and detects the troughs using the local maximum filter.
    Returns a boolean mask of the troughs (i.e. 1 when
    the pixel's value is the neighborhood maximum, 0 otherwise)
    """
    # define an connected neighborhood
    # http://www.scipy.org/doc/api_docs/SciPy.ndimage.morphology.html#generate_binary_structure
    neighborhood = morphology.generate_binary_structure(len(arr.shape),2)
    # apply the local minimum filter; all locations of minimum value 
    # in their neighborhood are set to 1
    # http://www.scipy.org/doc/api_docs/SciPy.ndimage.filters.html#minimum_filter
    local_min = (filters.minimum_filter(arr, footprint=neighborhood)==arr)
    # local_min is a mask that contains the peaks we are 
    # looking for, but also the background.
    # In order to isolate the peaks we must remove the background from the mask.
    # 
    # we create the mask of the background
    background = (arr==0)
    # 
    # a little technicality: we must erode the background in order to 
    # successfully subtract it from local_min, otherwise a line will 
    # appear along the background border (artifact of the local minimum filter)
    # http://www.scipy.org/doc/api_docs/SciPy.ndimage.morphology.html#binary_erosion
    eroded_background = morphology.binary_erosion(
        background, structure=neighborhood, border_value=1)
    # 
    # we obtain the final mask, containing only peaks, 
    # by removing the background from the local_min mask
    detected_minima = local_min ^ eroded_background
    return np.where(detected_minima)

可以像这样使用:

arr=np.array([[[0,0,0,-1],[0,0,0,0],[0,0,0,0],[0,0,0,0],[-1,0,0,0]],
              [[0,0,0,0],[0,-1,0,0],[0,0,0,0],[0,0,0,-1],[0,0,0,0]]])
local_minima_locations = detect_local_minima(arr)
print(arr)
# [[[ 0  0  0 -1]
#   [ 0  0  0  0]
#   [ 0  0  0  0]
#   [ 0  0  0  0]
#   [-1  0  0  0]]
#  [[ 0  0  0  0]
#   [ 0 -1  0  0]
#   [ 0  0  0  0]
#   [ 0  0  0 -1]
#   [ 0  0  0  0]]]

这表示最小值出现的索引为[0,0,3]，[0,4,0]，[1,1,1]和[1,3,3]:

print(local_minima_locations)
# (array([0, 0, 1, 1]), array([0, 4, 1, 3]), array([3, 0, 1, 3]))
print(arr[local_minima_locations])
# [-1 -1 -1 -1]

0