use fdmath derivatives where possible
This commit is contained in:
parent
e69b0f61ec
commit
e5db39fd49
@ -2,8 +2,7 @@
|
||||
from typing import List, Callable, Tuple, Dict
|
||||
import numpy
|
||||
|
||||
from .. import dx_lists_t, field_t, field_updater
|
||||
|
||||
from .. import dx_lists_t, field_t, field_updater, fdmath
|
||||
|
||||
def poynting(e: field_t,
|
||||
h: field_t,
|
||||
@ -35,10 +34,8 @@ def poynting_divergence(s: field_t = None,
|
||||
if s is None:
|
||||
s = poynting(e, h, dxes=dxes)
|
||||
|
||||
#TODO use deriv operators
|
||||
ds = ((s[0] - numpy.roll(s[0], 1, axis=0)) +
|
||||
(s[1] - numpy.roll(s[1], 1, axis=1)) +
|
||||
(s[2] - numpy.roll(s[2], 1, axis=2)))
|
||||
Dx, Dy, Dz = fdmath.functional.deriv_back()
|
||||
ds = Dx(s[0]) + Dy(s[1]) + Dz(s[2])
|
||||
return ds
|
||||
|
||||
|
||||
|
Loading…
Reference in New Issue
Block a user