For GrB_Vector_select, the intermediate vector t is in the same domain as the input vector u, which is correct.
For GrB_Matrix_select, the intermediate matrix T is said to be in the domain of the output of the operator, which is incorrect. The domain of T should be the same as the domain of the input matrix A.
For GrB_Vector_select, the intermediate vector t is in the same domain as the input vector u, which is correct.
For GrB_Matrix_select, the intermediate matrix T is said to be in the domain of the output of the operator, which is incorrect. The domain of T should be the same as the domain of the input matrix A.
See line 6670 of the PDF, as of Sept 10 at least.