Thanks for your code.It's amazing， I ran this code with a big data which size is 35003500. It cost lease than 1 second. But I have an question about the input data. The code only supports square matrix， such as 33 , 66 or 5050. But It doesn't support rectangular matrix, such as 34 , 58 , or 50*80. I wonder my way to solve it is right or not. When I have an rectangular matrix, for example:
b = np.array(
[[3,4,6,4,9,5],
[6,4,5,3,8,3],
[6,3,2,2,5,3],
[8,4,5,4,7,2]]
)
This is an rectangular matrix :4 * 6, so I fill it with an big number,:
And then, I can run the code ,is it right ?
And given a matrix of positive weights, the code finds the minimum weight bipartite matching between to arrays. If I want to find the max， I only need to use the max element of the matrix minus every element. So， if
someone element unmatched rows in the case of unbalanced entries，what index will return ？ -1 ？ or ？
Thanks for your code.It's amazing， I ran this code with a big data which size is 35003500. It cost lease than 1 second. But I have an question about the input data. The code only supports square matrix， such as 33 , 66 or 5050. But It doesn't support rectangular matrix, such as 34 , 58 , or 50*80. I wonder my way to solve it is right or not. When I have an rectangular matrix, for example:
This is an rectangular matrix :4 * 6, so I fill it with an big number,:
And then, I can run the code ,is it right ? And given a matrix of positive weights, the code finds the minimum weight bipartite matching between to arrays. If I want to find the max， I only need to use the max element of the matrix minus every element. So， if someone element unmatched rows in the case of unbalanced entries，what index will return ？ -1 ？ or ？