Describe the bug
The laplacian operator takes as input a method argument, which allows to choose how to compute the operator. Accepted possibilities are "std" or "divgrad".
This value is currently ignored as it is not forwarded to fast_laplacian, which uses the default setting.