Closed stigrj closed 2 years ago
Merging #400 (a565cc7) into master (ab262e1) will decrease coverage by
0.51%
. The diff coverage is0.82%
.:exclamation: Current head a565cc7 differs from pull request most recent head 5a5d2cb. Consider uploading reports for the commit 5a5d2cb to get more accurate results
@@ Coverage Diff @@
## master #400 +/- ##
==========================================
- Coverage 68.51% 68.00% -0.52%
==========================================
Files 179 179
Lines 14300 14410 +110
==========================================
+ Hits 9798 9799 +1
- Misses 4502 4611 +109
Impacted Files | Coverage Δ | |
---|---|---|
src/initial_guess/gto.cpp | 59.01% <0.00%> (-19.25%) |
:arrow_down: |
src/initial_guess/sad.cpp | 67.93% <0.00%> (-32.07%) |
:arrow_down: |
src/qmfunctions/qmfunction_utils.cpp | 81.02% <0.00%> (-5.79%) |
:arrow_down: |
src/driver.cpp | 69.55% <50.00%> (-0.06%) |
:arrow_down: |
src/chemistry/Nucleus.h | 73.91% <0.00%> (-26.09%) |
:arrow_down: |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update ab262e1...5a5d2cb. Read the comment docs.
SAD_GTO
convergence for some Ni complex:
===========================================================================
Iter MO residual Total energy Update
---------------------------------------------------------------------------
0 9.486833e+00 -2871.133370005639 -2.871133e+03
1 6.373926e-01 -2886.516418844161 -1.538305e+01
2 2.557858e-01 -2887.134570848163 -6.181520e-01
3 1.194583e-01 -2887.245656006424 -1.110852e-01
4 7.999071e-02 -2887.286245156482 -4.058915e-02
5 1.264002e-01 -2887.321871333240 -3.562618e-02
6 6.671218e-02 -2887.334927415342 -1.305608e-02
7 6.802234e-02 -2887.347317597009 -1.239018e-02
8 4.746045e-02 -2887.353855939271 -6.538342e-03
9 5.211014e-02 -2887.357988889306 -4.132950e-03
10 2.794708e-02 -2887.360939947434 -2.951058e-03
11 2.283775e-02 -2887.362321669932 -1.381722e-03
12 1.505338e-02 -2887.362824301893 -5.026320e-04
---------------------------------------------------------------------------
SCF converged in 12 iterations!
===========================================================================
compared to the good ol' SAD_DZ
:
===========================================================================
Iter MO residual Total energy Update
---------------------------------------------------------------------------
0 9.486833e+00 -2851.355144659785 -2.851355e+03
1 1.652298e+00 -2866.293443194611 -1.493830e+01
2 1.124413e+00 -2873.238534223135 -6.945091e+00
3 7.123881e-01 -2876.870289729848 -3.631756e+00
4 6.979813e-01 -2879.453487196854 -2.583197e+00
5 6.689936e-01 -2881.418573873416 -1.965087e+00
6 4.042748e-01 -2882.406062604978 -9.874887e-01
7 1.513471e+00 -2884.282844013223 -1.876781e+00
8 5.992928e-01 -2884.997568455024 -7.147244e-01
9 7.527686e-01 -2885.971568578544 -9.740001e-01
10 3.291256e-01 -2886.427363244316 -4.557947e-01
11 4.072887e-01 -2886.719450121097 -2.920869e-01
12 4.008856e-01 -2886.953903769770 -2.344536e-01
13 2.179256e-01 -2887.054556525391 -1.006528e-01
14 5.348428e-01 -2887.128757250646 -7.420073e-02
15 3.394773e-01 -2887.158681634962 -2.992438e-02
16 4.588629e-01 -2887.239357284327 -8.067565e-02
17 1.602679e-01 -2887.293899982501 -5.454270e-02
18 8.510131e-02 -2887.316638160011 -2.273818e-02
19 1.805406e-01 -2887.333276527359 -1.663837e-02
20 7.475968e-02 -2887.345227063257 -1.195054e-02
21 1.382609e-01 -2887.355325514161 -1.009845e-02
22 3.819932e-02 -2887.359147728878 -3.822215e-03
23 6.187910e-02 -2887.362356608575 -3.208880e-03
24 2.858246e-02 -2887.364267834016 -1.911225e-03
25 3.096303e-02 -2887.365610029450 -1.342195e-03
26 1.981330e-02 -2887.366463058896 -8.530294e-04
27 1.537261e-02 -2887.367044335408 -5.812765e-04
---------------------------------------------------------------------------
SCF converged in 27 iterations!
===========================================================================
Very nice! :)
Here is sad_gto
for Xe
(world_prec = guess_prec = 1e-5
)
Walltime: 18 minutes
===========================================================================
Iter MO residual Total energy Update
---------------------------------------------------------------------------
0 5.196152e+00 -7202.619249713141 -7.202619e+03
1 1.160183e-01 -7233.031488098271 -3.041224e+01
2 3.283643e-02 -7234.076552256704 -1.045064e+00
3 1.667258e-02 -7234.196645609136 -1.200934e-01
4 1.310087e-02 -7234.220586582349 -2.394097e-02
5 1.170140e-02 -7234.231655376288 -1.106879e-02
6 4.885828e-03 -7234.232218701658 -5.633254e-04
7 3.572509e-03 -7234.232818258200 -5.995565e-04
8 1.560652e-03 -7234.232917946327 -9.968813e-05
9 8.057152e-04 -7234.232964806587 -4.686026e-05
10 5.494256e-04 -7234.232990455595 -2.564901e-05
11 3.050460e-04 -7234.232993350029 -2.894434e-06
12 1.277010e-04 -7234.232994278396 -9.283676e-07
---------------------------------------------------------------------------
SCF converged in 12 iterations!
===========================================================================
and the same with sad_dz
Walltime: 25 minutes
===========================================================================
Iter MO residual Total energy Update
---------------------------------------------------------------------------
0 5.196152e+00 -7134.586380860554 -7.134586e+03
1 1.946345e+00 -7218.310867616008 -8.372449e+01
2 1.058071e+00 -7228.682418878908 -1.037155e+01
3 6.361629e-01 -7232.392158178065 -3.709739e+00
4 2.130290e-01 -7233.398960646703 -1.006802e+00
5 1.850029e-01 -7233.982924021013 -5.839634e-01
6 1.041522e-01 -7234.176739783486 -1.938158e-01
7 5.698223e-02 -7234.214943367646 -3.820358e-02
8 2.991435e-02 -7234.225195276479 -1.025191e-02
9 1.881444e-02 -7234.230164071142 -4.968795e-03
10 9.468968e-03 -7234.231917155496 -1.753084e-03
11 1.053953e-02 -7234.232192296029 -2.751405e-04
12 5.704532e-03 -7234.232711716407 -5.194204e-04
13 3.340135e-03 -7234.232850554931 -1.388385e-04
14 2.414959e-03 -7234.232918021183 -6.746625e-05
15 1.892349e-03 -7234.232948312133 -3.029095e-05
16 1.297262e-03 -7234.232964464264 -1.615213e-05
17 6.341829e-04 -7234.232969582814 -5.118550e-06
18 4.916449e-04 -7234.232971318601 -1.735787e-06
19 3.021556e-04 -7234.232972460484 -1.141883e-06
20 2.148755e-04 -7234.232973139436 -6.789514e-07
---------------------------------------------------------------------------
SCF converged in 20 iterations!
===========================================================================
New
SCF.guess_type = sad_gto
which reads the 3-21G basis sets fromshare/sad_basis/
instead of setting up an AO basis of Hydrogen eigenfunctions. Not sure whether it's an improvement over the standardsad_dz
, though...Added screening option for the evaluation of GTOs in various initial guesses. Controlled by input parameter
SCF.guess_screen
which is afloat
representing number of standard deviations, beyond which the GTO is evaluated to zero. Expected to speed up standard SAD guesses somewhat.