Skip to content

Conversation

thomas-saigre
Copy link
Contributor

@thomas-saigre thomas-saigre commented Jun 20, 2023

[ci skip]

  • Have you checked to ensure there aren't other open Pull Requests for the same change?
  • Have you added an explanation of what your changes do and why you'd like us to include them?
  • Have you written new tests for your changes?
  • Have you successfully run the Feel++ testsuite with your changes locally?
  • Have you written Doxygen comments in your contribution ?

During offline generation, if CRB mode is selected (run.mode=2), the effectivity
$\eta_N(\mu) = \dfrac{\Delta_N(\mu)}{|s(\mu) - s_N(\mu)|}$ is computed and displayed in the array of result :

mu0        mu1        mu2        mu3        mu4        mu5        mu6            FEM Output     FEM Time       RB Output      Error Bounds   effectivity    CRB Time       output error   Conditionning  l2_error       h1_error
3.0066e-01 4.7983e+01 9.3008e+01 6.0000e+00 1.0000e+00 1.8284e+04 2.9131e+04     3.2157e+02     1.9985e+01     3.2156e+02     1.5206e-04     6.7484e+00     5.6476e-05     2.2533e-05     0.0000e+00     7.9557e-04     3.7919e-01
4.1319e-01 7.1171e+01 6.1887e+01 6.0000e+00 1.0000e+00 3.2053e+04 2.6068e+04     4.1876e+02     1.9648e+01     4.1875e+02     1.3048e-04     6.2781e+00     5.0956e-05     2.0784e-05     0.0000e+00     7.8023e-04     5.8270e-01
2.1864e-01 7.0397e+01 6.3517e+01 6.0000e+00 1.0000e+00 5.0270e+03 1.5658e+04     1.7065e+02     1.9356e+01     1.7066e+02     2.1818e-04     6.4827e+00     5.7828e-05     3.3655e-05     0.0000e+00     1.4593e-03     6.9733e-02
5.2581e-01 4.6177e+01 8.2563e+01 6.0000e+00 1.0000e+00 9.3250e+03 3.0219e+04     3.0579e+02     1.9552e+01     3.0577e+02     4.3747e-04     6.4856e+00     5.8500e-05     6.7453e-05     0.0000e+00     1.0874e-03     9.3371e-02
3.5475e-01 3.5148e+01 8.6927e+01 6.0000e+00 1.0000e+00 2.9229e+04 1.6398e+04     3.3276e+02     2.0125e+01     3.3275e+02     1.8816e-04     5.9809e+00     5.6497e-05     3.1460e-05     0.0000e+00     1.2304e-03     2.6739e-02
4.8651e-01 6.4633e+01 6.9587e+01 6.0000e+00 1.0000e+00 2.8460e+04 3.3563e+04     4.5115e+02     1.9792e+01     4.5113e+02     2.1887e-04     6.3886e+00     7.3478e-05     3.4259e-05     0.0000e+00     7.7984e-04     1.6710e-01
2.3826e-01 8.8390e+01 8.1406e+01 6.0000e+00 1.0000e+00 2.3625e+04 2.4237e+04     2.7622e+02     1.9888e+01     2.7622e+02     5.2005e-05     7.7256e+00     5.2529e-05     6.7316e-06     0.0000e+00     3.3788e-04     6.9162e-02
2.3656e-01 3.0962e+01 8.1230e+01 6.0000e+00 1.0000e+00 5.4438e+03 3.2835e+04     3.3703e+02     2.0772e+01     3.3703e+02     1.8646e-05     2.6001e+01     5.5715e-05     7.1715e-07     0.0000e+00     8.8732e-05     6.6041e-03
2.5166e-01 8.7458e+01 1.0170e+02 6.0000e+00 1.0000e+00 1.7226e+04 2.0708e+04     1.9511e+02     1.8827e+01     1.9511e+02     8.3012e-05     9.9277e+00     5.9262e-05     8.3616e-06     0.0000e+00     5.5793e-04     1.8885e-01
3.0750e-01 4.2566e+01 9.0083e+01 6.0000e+00 1.0000e+00 2.2389e+04 2.1830e+04     3.0962e+02     1.9383e+01     3.0961e+02     2.1045e-04     6.4528e+00     6.3911e-05     3.2613e-05     0.0000e+00     1.2545e-03     7.1128e-02

NB : The effectityvy is obtained from the absolute error, not the relative one.

thomas-saigre and others added 5 commits May 23, 2023 10:11
(not working yet) #2098

[ci skip]
refactor constructor of OpusApp class

[skip feelpp, skip tests, skip toolboxes, skip python]
Co-authored-by: Christophe Prud'homme <christophe.prudhomme@cemosis.fr>
refactor opusapp_impl : remove duplication of code
#2098
[skip feelpp, skip toolboxes, skip tests, skip python]
/cc @prudhomm
@thomas-saigre thomas-saigre linked an issue Jun 20, 2023 that may be closed by this pull request
@thomas-saigre
Copy link
Contributor Author

@prudhomm the runner did pass well

@prudhomm prudhomm merged commit 3708ba2 into develop Jun 25, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
Status: Done
Development

Successfully merging this pull request may close these issues.

Compute effectivity
2 participants