Why are rows and column references transposed in mojo matrix implementations?

[cheburakshu]

Source: https://docs.modular.com/mojo/notebooks/Matmul.html#adding-types-to-the-python-implementation

fn matmul_naive(C: Matrix, A: Matrix, B: Matrix):
    for m in range(C.rows):
        for k in range(A.cols):
            for n in range(C.cols):
                C[m, n] += A[m, k] * B[k, n]

A[m, k] in the above code appears to reference index m for row and k for column, whereas in the __getitem__ method, it is transposed. There are other examples in documentation where this is happening too, but is there any specific reason the implementation is transposing row and column references?

    @always_inline
    fn __getitem__(self, y: Int, x: Int) -> Float32:
        return self.load[1](y, x)

[gdonval]

That's a typical case of bad naming: it's not transposed. y is the row index. If you have a matrix like that:

[[a, b, c],
 [d, e, f]]

there are 2 rows and 3 columns. d is on [1, 0], i.e. y=1 and x=0 so 1 * 3 + 0 = 3, which is the correct index (in 0-based indexing).

Getitem is defined with y, x, not the other way around.