ABINIT Discussion Forums

Dear All,

I am quite new in the community.

I have successfully compiled abinit8.0.8 on our debian8 plateform.

After executing the tests, I observed some deviations to the expected values (quite large from my point of view).

Below a summary of the execution of tests and attached is my config.log file:

make test_fast
cp ./src/98_main/abinit ../tests/built-in/Input/abinit && cd ../tests/built-in/Input && ./abinit < testin_fast.files >& testin_fast.stdout && cat testin_fast_tmp_STATUS && rm -f abinit *DDB *EIG *out* *nc *WFK

Status file, reporting on built-in test fast

==> The run finished cleanly.
Moreover, comparison of the total energy, and other (few) relevant quantities with reference values has been successful.
This does not mean that no problem is present, however.
Please run the complete set of ABINIT tests to gain a better confidence in your installation.

make test_v1
cp ./src/98_main/abinit ../tests/built-in/Input/abinit && cd ../tests/built-in/Input && ./abinit < testin_v1.files >& testin_v1.stdout && cat testin_v1_tmp_STATUS && rm -f abinit *DDB *EIG *out* *nc *WFK

Status file, reporting on built-in test v1

==> The run finished cleanly.
Moreover, comparison of the total energy, and other (few) relevant quantities with reference values has been successful.
This does not mean that no problem is present, however.
Please run the complete set of ABINIT tests to gain a better confidence in your installation.

make test_v5
cp ./src/98_main/abinit ../tests/built-in/Input/abinit && cd ../tests/built-in/Input && ./abinit < testin_v5.files >& testin_v5.stdout && cat testin_v5_tmp_STATUS && rm -f abinit *DDB *EIG *out* *nc *WFK

Status file, reporting on built-in test v5

==> The run finished cleanly.
Moreover, comparison of the total energy, and other (few) relevant quantities with reference values has been successful.
This does not mean that no problem is present, however.
Please run the complete set of ABINIT tests to gain a better confidence in your installation.

make test_bigdft
cp ./src/98_main/abinit ../tests/built-in/Input/abinit && cd ../tests/built-in/Input && ./abinit < testin_bigdft.files >& testin_bigdft.stdout && cat testin_bigdft_tmp_STATUS && rm -f abinit *DDB *EIG *out* *nc *WFK

Status file, reporting on built-in test bigdft

Error for total energy :
expected -0.56231990141000D+00 with maximum deviation 0.3000D-06
computed -0.56217058543260D+00 with effective deviation 0.1493D-03
Error for stress strten(1)
expected -0.22537238382000D-01 with maximum deviation 0.1000D-06
computed -0.22369007705030D-01 with effective deviation 0.1682D-03
Error for stress strten(2)
expected -0.22536232141000D-01 with maximum deviation 0.1000D-06
computed -0.22364516784783D-01 with effective deviation 0.1717D-03
Error for stress strten(3)
expected -0.22529038043000D-01 with maximum deviation 0.1000D-06
computed -0.22359905383017D-01 with effective deviation 0.1691D-03
Error for stress strten(4)
expected 0.39104928264000D-06 with maximum deviation 0.1000D-06
computed 0.17586904149594D-06 with effective deviation 0.2152D-06
Error for stress strten(5)
expected 0.62823846408000D-06 with maximum deviation 0.1000D-06
computed 0.29437702374832D-06 with effective deviation 0.3339D-06

make test_etsf_io
cp ./src/98_main/abinit ../tests/built-in/Input/abinit && cd ../tests/built-in/Input && ./abinit < testin_etsf_io.files >& testin_etsf_io.stdout && cat testin_etsf_io_tmp_STATUS && rm -f abinit *DDB *EIG *out* *nc *WFK

Status file, reporting on built-in test etsf_io

Error for force fred(2,1)
expected -0.12087700000000D-06 with maximum deviation 0.1000D-06
computed -0.49569465260213D-09 with effective deviation 0.1204D-06
Error for force fred(3,1)
expected -0.16428700000000D-06 with maximum deviation 0.1000D-06
computed 0.17659226101742D-07 with effective deviation 0.1819D-06

make test_libxc
cp ./src/98_main/abinit ../tests/built-in/Input/abinit && cd ../tests/built-in/Input && ./abinit < testin_libxc.files >& testin_libxc.stdout && cat testin_libxc_tmp_STATUS && rm -f abinit *DDB *EIG *out* *nc *WFK

Status file, reporting on built-in test libxc

Error for total energy :
expected -0.55196688523897D+01 with maximum deviation 0.5000D-06
computed -0.55188626213103D+01 with effective deviation 0.8062D-03
Error for stress strten(1)
expected 0.13246699375127D-04 with maximum deviation 0.1000D-07
computed 0.17283635701844D-04 with effective deviation 0.4037D-05
Error for stress strten(2)
expected 0.13246699375127D-04 with maximum deviation 0.1000D-07
computed 0.17283635701844D-04 with effective deviation 0.4037D-05
Error for stress strten(3)
expected 0.13246699375127D-04 with maximum deviation 0.1000D-07
computed 0.17283635701844D-04 with effective deviation 0.4037D-05

make test_wannier90
cp ./src/98_main/abinit ../tests/built-in/Input/abinit && cd ../tests/built-in/Input && ./abinit < testin_wannier90.files >& testin_wannier90.stdout && cat testin_wannier90_tmp_STATUS && rm -f abinit *DDB *EIG *out* *nc *WFK *DEN *chk *eig *mmn *amn

Status file, reporting on built-in test wannier90

Error for total energy :
expected -0.10620085133544D+02 with maximum deviation 0.3000D-06
computed -0.10619010196886D+02 with effective deviation 0.1075D-02
Error for stress strten(1)
expected 0.31413922197317D-03 with maximum deviation 0.1000D-05
computed 0.98164613957580D-04 with effective deviation 0.2160D-03
Error for stress strten(2)
expected 0.31413922197317D-03 with maximum deviation 0.1000D-05
computed 0.98164613957580D-04 with effective deviation 0.2160D-03
Error for stress strten(3)
expected 0.31413922197317D-03 with maximum deviation 0.1000D-05
computed 0.98164613957580D-04 with effective deviation 0.2160D-03

Is it normal to have such deviations, even for the total energy

Thanks in advance

Rémi

Hi,

Depending on you cpu, your librairies, and so one, deviations are expected. It is hard to say if they're significant or not in you case since they are quite larger than the deviation extracted from our test farm.
This being said, our test farm is not exhaustive so the expected deviation may be wrong.
To be sure, the developers of the features you tested and did not passed should have a look but we may not be able to reproduce your results. So it is very hard to debug.
You can run the full testfarm to have a larger overview of what seems to work and does not.

Cheers