uniform_move_snippet.f90
· 895 B · Fortran
Eredeti
! For the polar angle part, since projection is involved and therefore
! we need to guarantee that the projection is uniform. We don't need
! to do this for the phi angle since only rotation is involved in
! that case.
polar_ang = dacos(2.0d0*(rmca_ran(1)-0.5d0))
azim_ang = rmca_ran(1) * 2.0d0 * pi
rand_x = delta(move_type) * dsin(polar_ang) * dcos(azim_ang)
rand_y = delta(move_type) * dsin(polar_ang) * dsin(azim_ang)
rand_z = delta(move_type) * dcos(polar_ang)
! We go from the Cartesian coordinates to fractional with respect
! to the lattice.
move(4) = rand_x * inv_vectors_init(1,1) + &
rand_y * inv_vectors_init(1,2) + &
rand_z * inv_vectors_init(1,3)
move(5) = rand_x * inv_vectors_init(2,1) + &
rand_y * inv_vectors_init(2,2) + &
rand_z * inv_vectors_init(2,3)
move(6) = rand_x * inv_vectors_init(3,1) + &
rand_y * inv_vectors_init(3,2) + &
rand_z * inv_vectors_init(3,3)
1 | ! For the polar angle part, since projection is involved and therefore |
2 | ! we need to guarantee that the projection is uniform. We don't need |
3 | ! to do this for the phi angle since only rotation is involved in |
4 | ! that case. |
5 | polar_ang = dacos(2.0d0*(rmca_ran(1)-0.5d0)) |
6 | azim_ang = rmca_ran(1) * 2.0d0 * pi |
7 | rand_x = delta(move_type) * dsin(polar_ang) * dcos(azim_ang) |
8 | rand_y = delta(move_type) * dsin(polar_ang) * dsin(azim_ang) |
9 | rand_z = delta(move_type) * dcos(polar_ang) |
10 | ! We go from the Cartesian coordinates to fractional with respect |
11 | ! to the lattice. |
12 | move(4) = rand_x * inv_vectors_init(1,1) + & |
13 | rand_y * inv_vectors_init(1,2) + & |
14 | rand_z * inv_vectors_init(1,3) |
15 | move(5) = rand_x * inv_vectors_init(2,1) + & |
16 | rand_y * inv_vectors_init(2,2) + & |
17 | rand_z * inv_vectors_init(2,3) |
18 | move(6) = rand_x * inv_vectors_init(3,1) + & |
19 | rand_y * inv_vectors_init(3,2) + & |
20 | rand_z * inv_vectors_init(3,3) |
21 |