thatguynoe
bdf6aa6369
ensure root search interval is valid
...
We now look for a root of the function f(c) = ∫_Ω sigmoid(ψ + c) dx - θ vol(Ω) within the interval [a,b], where a := -‖αG‖_∞ and b := ‖αG‖_∞, and α and G are as in Step 4 and Step 5. It follows that f(a) ≤ ∫_Ω sigmoid(ψ + αG) dx - θ vol(Ω) ≤ f(b), and the inner quantity equals 0 since ψ_new := ψ_prev - αG in Step 5 and ∫_Ω sigmoid(ψ_new) dx - θ vol(Ω) = 0. This ensures f(a) ≤ 0 ≤ f(b), as required by the Illinois method.
2026-04-07 19:39:30 -04:00
thatguynoe
f6979648e8
move boundary assembly into bilinear form assembly
...
Prevents the user from accidentally calling these methods in the wrong order.
2025-05-30 09:17:52 -07:00
thatguynoe
2a4decc635
use the bisection method instead of Newton
2025-05-29 12:33:49 -07:00
thatguynoe
5808fc6966
use gradient descent step length in proj
...
See https://github.com/mfem/mfem/pull/4645#discussion_r1898110950 .
2025-02-25 19:21:07 -05:00
thatguynoe
b40bf6a64d
Revert "add option to choose bisection method for roots"
...
This reverts commit 21778cd9335337a1419af482a7feaf6baac1057f.
2025-02-25 18:31:42 -05:00
thatguynoe
7a73e97922
Revert "rename newton arg for clarity"
...
This reverts commit d7681c26dc608bfab1ca4c891e22cbc1260340c6.
2025-02-25 18:31:42 -05:00
thatguynoe
3a97122e34
rename newton arg for clarity
2025-02-25 18:31:42 -05:00
thatguynoe
2f89a16314
update sample runs, fix stability
2025-02-25 18:31:42 -05:00
thatguynoe
85fb20a1d1
set smaller itol for better convergence
2025-02-25 18:31:42 -05:00
thatguynoe
428d203eac
add option to choose bisection method for roots
2025-02-25 18:31:42 -05:00
thatguynoe
2e8f6f9c28
assemble boundary, bilinear form outside of solve
2025-02-25 18:31:42 -05:00
thatguynoe
11e4c46f25
add growth rate arg for grad descent step length
2025-02-25 18:31:42 -05:00
thatguynoe
ff8d8752c7
move proj function into header file
2025-02-25 18:31:42 -05:00
Dylan Copeland
046a874cfb
Merge branch 'master' of github.com:mfem/mfem into float
2024-02-06 10:56:41 -08:00
Aaron Fisher
7bbab61e2b
Significant improvements to the amgx solver doxygen. Fixed a couple doxygen errors in ex37.
2023-12-20 12:08:10 -08:00
Tzanio Kolev
d983b34924
Renamed ftype -> real_t
2023-12-13 12:51:43 -08:00
Dylan Copeland
b89222762d
Generalized floating point type for all remaining examples.
2023-11-06 15:34:42 -08:00
Tzanio Kolev
afcca4036c
Clean-up pass through new examples and miniapps
2023-09-26 08:36:59 -07:00
Brendan Keith and Veselin Dobrev
a3f0f89512
Update examples/ex37p.cpp
...
Co-authored-by: Veselin Dobrev <v-dobrev@users.noreply.github.com >
2023-09-19 14:25:12 -04:00
Brendan Keith
b5888d4b0f
add incremental exit tolerance, update comments, make variable and functions names align more closely with the paper
2023-08-30 22:56:04 -04:00
Brendan Keith
b5ceaca56c
minor
2023-08-29 13:59:48 -04:00
Brendan Keith
329cfb998c
Update comments in the code. Estimate error in the L1 norm
2023-08-29 13:57:28 -04:00
Brendan Keith
fc8477a265
remove clipping function, update sample runs, and change default mesh density
2023-08-23 18:20:02 -04:00
Brendan Keith
cda099923d
added paraview options
2023-08-21 11:34:06 -04:00
Brendan Keith
7bec37fe5b
adjust glvis output and comments
2023-08-11 17:46:48 -04:00
Brendan Keith
37a7e5466b
fix comments in ex37
2023-08-11 16:29:41 -04:00
Brendan Keith
be168f4b78
fixing white space mistake
2023-08-11 16:24:55 -04:00
Brendan Keith
99af756462
update files
2023-08-11 16:12:42 -04:00
Brendan Keith
d68a188884
rename examples
2023-08-10 21:58:28 -04:00
thatguynoe