Conversation
rouson
left a comment
rouson
left a comment
There was a problem hiding this comment.
In addition to the requested changes, please also either remove commented code or make the code conditional so that setting some parameter will actually execute it. In an ideal world, there would also be a test that sets that parameter to make sure the code actually works.
src/matcha/grid_m.f90
Outdated
| integer, intent(in) :: ndim | ||
| integer, intent(in) :: ng(:) | ||
| double precision, intent(in) :: gb,ge | ||
| double precision gridp(2*ndim+1) |
There was a problem hiding this comment.
@davytorres because gridp is not a dummy argument, please move it to the grid_s submodule, make it allocatable and then allocate it there.
| double precision gridp(2*ndim+1) | ||
| end function gridparameters | ||
|
|
||
| module function gradient(self, my_num_cells, ng, dx, gb, tconc, x, concentration_subgrid) result(gx) |
There was a problem hiding this comment.
@davytorres it appears likely that this procedure could be pure.
There was a problem hiding this comment.
Because concentration_subgrid is modified inside the procedure, making gradient pure will also require converting it to a subroutine and giving concentration_subgrid the attribute intent(inout).
There was a problem hiding this comment.
I tried making the changes but the co_sum call cannot be used in a pure procedure.
| interface | ||
|
|
||
| module function simulated_distribution(self) result(output_distribution) | ||
| !result(output_distribution,output_angle_distribution) |
There was a problem hiding this comment.
Please remove commented code.
| end function | ||
|
|
||
| module function simulated_distribution_angles(self) result(output_angle_distribution) | ||
| !result(output_distribution,output_angle_distribution) |
There was a problem hiding this comment.
Please remove commented code.
3 participants
No description provided.