Convergence of the QP direct gaps (eV) at Gamma and X with the number of points in the minimax mesh.
The color of each point is proportional to the duality error Max_{ij} |CT CT^{-1} - I|.
Convergence of the RPA energy (eV) with the number of points in the minimax mesh.
The color of each point is proportional to the duality error Max_{ij} |CT CT^{-1} - I|.
Note that the results obtained with ntau == 6 are not shown.
FIXME Note that I'm rescaling the results obtained with the last 5 meshes (ntau 26-34) by the infamous factor 4 (spin and integration range). Very likely the weights for the last 5 meshes have been generated using a different convention. This is what I get for LiF if I don't rescale
ntau rpa_ec_ev ft_max_err_t2w_cos ft_max_err_w2t_cos ft_max_err_t2w_sin cosft_duality_error
0 6 -14.499451 1.933901e-03 1.353624e-03 4.325153e-02 1.309390e-03
1 8 -14.499090 4.909689e-04 1.179800e-04 9.218434e-03 2.711183e-03
2 10 -14.499092 5.332045e-05 3.952615e-05 1.896179e-03 1.762042e-03
3 12 -14.499092 2.246161e-05 2.504123e-06 3.800066e-04 8.050877e-03
4 14 -14.499092 4.629865e-06 4.686459e-07 7.477280e-05 1.314336e-02
5 16 -14.499092 9.073634e-07 8.641273e-08 1.448701e-05 2.097865e-02
6 18 -14.499093 6.110221e-07 1.125080e-07 2.764762e-06 3.113177e+00
7 20 -14.499094 1.671891e-06 1.810501e-07 2.255550e-06 5.147856e-01
8 22 -14.499093 4.704446e-08 3.180201e-08 1.027178e-07 1.528265e+02
9 24 -14.499093 1.962311e-08 1.929468e-08 1.914650e-07 5.472928e+07
10 26 -3.624773 7.389984e-09 8.245343e-09 7.419598e-08 1.602566e+06
11 28 -3.624773 1.088852e-08 8.016416e-09 1.227601e-07 1.232216e+06
12 30 -3.624773 1.109270e-08 7.893361e-09 1.227774e-07 1.206735e+06
13 32 -3.624773 1.037138e-08 9.264811e-09 1.311225e-07 1.986793e+06
14 34 -3.624773 1.325711e-08 1.351992e-08 1.575436e-07 2.962972e+06
