按行比较两个大小不等但维度相同的数组。 （Python）

Question

I have two arrays A and B of sizes (m,2) and (n,2). A and B store the indices of non-zero elements in two different matrices P and Q.

I want to find the indices which match, so I can find which indices store non-zero values.

Answer 1

Not 100% certain that this is what you're looking for, as you haven't given an example input and output, but here's what I think you want. Essentially, looping through each element in the first array and searching if that element exists in the second array. Then flattening the output array.

a = [[1,2],[3,4],[5,6],[7,8],[9,10],[11,12]]
b = [[5,7],[2,9],[13,12]]

np.shape(a)  #(6,2)
np.shape(b)  #(3,2)

# Find items common to both
c = [[[item for item in a[i] if item in b[j]] for i in range(len(a))] for j in range(len(b))]

# [[[], [], [5], [7], [], []],
#  [[2], [], [], [], [9], []],
#  [[], [], [], [], [], [12]]]

# Remove empties and flatten
c = [item for elements in c for items in elements for item in items if item != []]
#  [5, 7, 2, 9, 12]

EDIT:

Possibly a more reader-friendly method is to first flatten a and b:

a_f = [item for items in a for item in items]
# [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12]

b_f = [item for items in b for item in items]
# [5, 7, 2, 9, 13, 12]

Then perform the loop through:

c = [item for item in a_f if item in b_f]
# [2, 5, 7, 9, 12]

EDIT2:

I may have misunderstood the question. If the elemnts in the arrays are two-dimensional coordinates/indicies, then you just need this:

a = [[1,2],[3,4],[5,6],[7,8],[9,10],[11,12]]
b = [[5,6],[1,14],[8,7],[3,4]]

c = [item for item in a if item in b]
#  [[3, 4], [5, 6]]