Relax lda_u/02-ACBN0.test tolerances

Description

The tolerance for the lda_u/02-ACBN0.test seems to be too tight.

Failed tests

lda_u/02-ACBN0.01-nio.inp
-----------------------------------------
Match  Hubbard energy      :

   Calculated value : 0.14309432
   Reference value  : 0.143092905
   Difference       : 1.41500000000461e-06
   Deviation [%]    : 0.000988868036472254
   Tolerance        : 3.02e-07
   Tolerance [%]    : 0.000211051694002578

  Hubbard energy                        :        [  FAIL  ]
-----------------------------------------
Match  Local Magnetic Moment (Ni1)  :

   Calculated value : 1.878218
   Reference value  : 1.8782205
   Difference       : 2.49999999990536e-06
   Deviation [%]    : 0.000133104712673797
   Tolerance        : 1.65e-06
   Tolerance [%]    : 8.78491103680319e-05

  Local Magnetic Moment (Ni1)           :        [  FAIL  ]
-----------------------------------------
Match  Local Magnetic Moment (Ni2)  :

   Calculated value : -1.878218
   Reference value  : -1.8782205
   Difference       : 2.49999999990536e-06
   Deviation [%]    : 0.000133104712673797
   Tolerance        : 1.65e-06
   Tolerance [%]    : 8.78491103680319e-05

  Local Magnetic Moment (Ni2)           :        [  FAIL  ]
-----------------------------------------
Match  Eigenvalue 16  :

   Calculated value : -0.822704
   Reference value  : -0.822708
   Difference       : 4.000000000004e-06
   Deviation [%]    : 0.000486199234722891
   Tolerance        : 3.3e-06
   Tolerance [%]    : 0.000401114368645984

  Eigenvalue 16                         :        [  FAIL  ]
-----------------------------------------
Match  Ueff 3d Ni1  :

   Calculated value : 0.550576
   Reference value  : 0.550582
   Difference       : 6.00000000006151e-06
   Deviation [%]    : 0.00108975593100783
   Tolerance        : 1.1e-06
   Tolerance [%]    : 0.000199788587349387

  Ueff 3d Ni1                           :        [  FAIL  ]
-----------------------------------------
Match  Ueff 3d Ni2  :

   Calculated value : 0.550577
   Reference value  : 0.550582
   Difference       : 5.00000000003276e-06
   Deviation [%]    : 0.000908129942503161
   Tolerance        : 1.1e-06
   Tolerance [%]    : 0.000199788587349387

  Ueff 3d Ni2                           :        [  FAIL  ]
-----------------------------------------
Match  Kanamori Up Ni1  :

   Calculated value : 0.687105
   Reference value  : 0.6871115000000001
   Difference       : 6.5000000001314e-06
   Deviation [%]    : 0.000945989115322826
   Tolerance        : 3.85e-06
   Tolerance [%]    : 0.000560316629833732

  Kanamori Up Ni1                       :        [  FAIL  ]
-----------------------------------------
Match  Kanamori Up Ni2  :

   Calculated value : 0.687106
   Reference value  : 0.6871130000000001
   Difference       : 7.00000000009027e-06
   Deviation [%]    : 0.00101875528480618
   Tolerance        : 2.2e-06
   Tolerance [%]    : 0.000320180232363527

  Kanamori Up Ni2                       :        [  FAIL  ]
-----------------------------------------

Test environment

• CPU: Intel® Xeon® Platinum 8360Y
• Libraries: gcc/10, openmpi/4, mkl/2021.2, gsl/2.4, hdf5-serial/1.8.21, netcdf-serial/4.4.1, libxc/4.3.4, metis/5.1, parmetis/4.0, cgal/5.2, boost/1.74, cuda/11.2, openmpi_gpu/4
• Configuration options : maxdim3 openmp mpi cuda sse2 avx libxc_fxc libxc_kxc
• Optional libraries : cgal metis netcdf parmetis scalapack
• Architecture : x86_64
• C compiler : mpicc (gcc)
• C compiler flags : -O3 -march=native -g
• C++ compiler : mpicxx (g++)
• C++ compiler flags : -O3 -march=native -g -std=c++14
• Fortran compiler : mpif90 (gfortran) (GCC version 10.3.0)
• Fortran compiler flags : -O3 -march=native -g -Wall -Wno-unused-dummy-argument -Wno-c-binding-type -fallow-argument-mismatch -fallow-invalid-boz -finit-int

added Testsuite label

@meisam.tabriz The intention is good, but I'm afraid this MR will not be very useful. The problem is that we don't have the same setup in the test farm, so as soon as someone updates this test (which unfortunately happens very often) the problem will reappear.

@micael.oliveira You are tight. But what should be the solution then? Each time I run the tests on Raven (or if someone tries to compile and test their own version there) it will fail.

Maybe we should think about a database for gathering the numerical results of our test runs on various architecture+environment.

I'm not sure if the current test parameters are even suited for our testfarm: https://octopus-code.org/buildbot/#/builders/106/builds/2516

Codecov Report

Merging #1590 (f115b5a3) into develop (594b19a8) will not change coverage. The diff coverage is n/a.

@@           Coverage Diff            @@
##           develop    #1590   +/-   ##
========================================
  Coverage    66.63%   66.63%           
========================================
  Files          537      537           
  Lines        94920    94920           
========================================
  Hits         63246    63246           
  Misses       31674    31674           

---

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 594b19a...f115b5a. Read the comment docs.

assigned to @micael.oliveira

changed milestone to %12.0

approved this merge request

resolved all threads

mentioned in commit afb2d311

merged