比较数组中的值并计算范围内的差异 Numpy Python

Question

我正在尝试编写一个函数来检查一个值是否在val103.0, 58.8, 35, -47 的值范围内下降超过 100，值从 103 变为 -47，即 -150 也是 58.8, 35, -47值从 58 变为 -47，减少了 -105。因此，有两种情况是从初始值下降到等于或超过 -100 的值。它检查数组中的每个值是否有 -100 的下降，直到数组末尾。因此，如果它采用数字7.4，那么它将检查数组的其余部分是否为1180.9, 0.6, 103.0, 58.8, 35, -47, 47.2, 78.1, 37.8，如果值为1180.9，那么它将检查值0.6, 103.0, 58.8, 35, -47, 47.2, 78.1, 37.8。我怎么能用 numpy 做到这一点。

import numpy as np 

val = np.array([7.4, 1180.9, 0.6, 103.0, 58.8, 35, -47, 47.2, 78.1, 37.8])
val2 = np.array([46.5, 55.7, 7.0, 19.6, 7.6, 36.5, 34.7, 101.9, 179.7, 85.5])
val3 = np.array([120, 20, -80, -5.5])

differences = 100

def run(values):
    minimums = np.subtract.accumulate(values)
    print(f'Number it has been below or equal to {differences} is {values}')
    return minimums
    
print(run(val))
print(run(val2))
print(run(val3))

预期输出：

Number it has been below or equal to -100 is 3
Number it has been below or equal to -100 is 0
Number it has been below or equal to -100 is 2

Answer 1

import numpy as np

val = np.array([7.4, 1180.9, 0.6, 103.0, 58.8, 35, -47, 47.2, 78.1, 37.8]) 
val2 = np.array([46.5, 55.7, 7.0, 19.6, 7.6, 36.5, 34.7, 101.9, 179.7, 85.5]) 
val3 = np.array([120, 20, -80, -5.5]) 

def run( arr, limit ): 
    row_mins = np.triu( arr[None, : ] - arr[ :, None ] ).min(axis = 1 ) 
    return np.count_nonzero( row_mins <= -limit ) 

扩展：

def run( arr, limit ):
    # Uncomment the print statements to see what is happening.
    temp = arr[None, : ] - arr[ :, None ]  # difference of each element all elements
    # print( temp )
    temp1 = np.triu( temp )  # Keep the upper triangle, set lower to zero
    # print( temp1 )
    row_mins = temp1.min( axis = 1 )   # Minimum for each row
    # print( row_mins )
    return np.count_nonzero( row_mins <= -limit ) # Count how many row_mins are <= limit

结果：

run(val, 100)   # 3
run(val2, 100)  # 0
run(val3, 100)  # 2