Build two array from another array based on the values and a window size

Question

I have an array with a thousand rows and columns. The array has the number 1, greater than 1, and less than 1. I want to build two arrays from that with this way:

The most important part is the values which are less than 1. Then based on a window size (here is 7), the value greater than 1 before the values less than 1, should change to 1, and all of the other remaining are zero. For example, if a row is [1, 1, 1.2, 0.5, 1.9, 1, 1], the first array that I want is: [0, 0, 1, 0,0,0,0]
and the second array that I want are related to the values greater than 1 after the values less than 1. For this example, I want [0, 0, 0, 0, 1, 0,0].

Here is a simple exmaple:
array I have:

a = np.array([[1,1,1.01, 0.5, 0.5, 1.02, 1, 1,1,1.21, 0.5, 0.5, 1.22, 1.3], [1,1.4,1.01, 0.5, 0.5, 1.02, 1, 1,1,1.51, 0.5, 0.7, 1.22, 1]])



 a= array([[1.  , 1.  , 1.01, 0.5 , 0.5 , 1.02, 1.  , 1.  , 1.  , 1.21, 0.5 ,0.5 , 1.22, 1.3 ],
           [1.  , 1.4 , 1.01, 0.5 , 0.5 , 1.02, 1.  , 1.  , 1.  , 1.51, 0.5 ,0.7 , 1.22, 1.  ]])

Two array I want:

array([[0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0],
       [0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0]])

and

array([[0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 1],
       [0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1]])

Could you please help me with this? Thank you

Asked By: sadcow

||

Source

Answer 1

You can achieve this by using the convolve1d from scipy:

from scipy.ndimage import convolve1d

def generate_arrays(a):
    # create mask arrays
    mask1 = np.zeros_like(a)
    mask2 = np.zeros_like(a)

    # create kernel for convolution
    kernel = np.ones(7)

    # convolve kernel with a
    convolved = convolve1d(a, kernel, axis=1, mode='constant')

    # create mask for values less than 1
    mask1[a < 1] = convolved[a < 1] < 1
    mask2[a < 1] = 0

    # create mask for values greater than 1
    mask1[a > 1] = 0
    mask2[a > 1] = convolved[a > 1] > 1

    return mask2, mask1

Answered By: DHJ

Build two array from another array based on the values and a window size

Question:

Answers: