search

sharc-md / sharc

The SHARC molecular dynamics (MD) program suite is an ab initio MD software package developed to study the excited-state dynamics of molecules.
https://www.sharc-md.org
GNU General Public License v3.0
63 stars 34 forks source link

SHARC-Openmolcas test job failure #131

Open maheshgiitdh opened 1 month ago

maheshgiitdh commented 1 month ago

Dear SHARC developers,

I recently installed SHARC and tried to run some test jobs with OPENMOLCAS interface. But all of them gave error as "Job did not finish successfully. Error code: 1".

The error in output.log file is " QM call was not successful, aborting the run Error code: 21"

I inspected all the molcas.output files in scratch directory, which are completed without any error and "happy landing"

At the end of QM.log corresponding to MOLCAS_DDR job, for instance, following errors have detected.

Starting the job execution START: /iitdh/faculty/maheshg/SCR/WORK/master 2024-07-10 18:29:13.126840 "/iitdh/faculty/maheshg/Programs/Molcas/Install/bin/pymolcas MOLCAS.input" FINISH: /iitdh/faculty/maheshg/SCR/WORK/master 2024-07-10 18:29:33.611353 Runtime: 0:00:20.484513 Error Code: 0

======================================== All Tasks completed!

========================================

Error Codes:

master          0       grad_1_1        0       grad_1_2        0       grad_3_1        0
    nacdr_1_1_1_2   0

Reading /iitdh/faculty/maheshg/SCR/WORK/master/MOLCAS.out DM element not found! 1 1 1 0 Found correct RASSI run, but too few matrix elements!

I am having the above error with the latest versions and OPENMOLCAS and SHARC.

Here, I have also attached the .out and .log files corresponding to the MOLCAS_DDR test job.

output.log QM.log MOLCAS_grad_1_1.txt MOLCAS_grad_1_2.txt MOLCAS_grad_3_1.txt MOLCAS_master.txt MOLCAS_nacdr.txt

Could you please help me to solve the issue.

Thank you very much for your time.

Mahesh

maisebastian commented 1 month ago

Dear Mahesh, thanks for the detailed report. It seems that in your version of OpenMolcas, the format in RASSI is minimally different, the dipole moments are given as PROPERTY: MLTPL 1 COMPONENT: 1 whereas in our installations it is PROPERTY: MLTPL 1 COMPONENT:* 1. The missing little asterisk in your output stopped the interface from finding the data.

Can you please download https://github.com/sharc-md/sharc/blob/sharc3preview/bin/SHARC_MOLCAS.py and try whether it fixes your problem?

Best, Sebastian

maheshgiitdh commented 1 month ago

Dear Sebastian,

Thank you very much for your prompt response.

I tried to run MOLCAS test jobs with the suggested SHARC_MOLCAS.py. It is still giving errors for some test jobs. Please find the summary of test jobs with error codes in the attached image.

MOLCAS_ddr and MOLCAS_overlap seem successfully finished, but the summary displays "Different output length"

MOLCAS_dyson gave similar error (QM call was not successful, aborting the run. Error code: 89) as mentioned (for MOLCAS_ddr) in my first email.

For MOLCAS_tCSDM_SA-CASSCF and MOLCAS_tCSDM_SA-CASSCF_with_TDH_scheme, the job quickly exits, without initiating the QM calculation, with an error code 127. The SCRATCH area also does not have any QM output files corresponding to these test jobs.

I have attached a zip file having the .out or .log files corresponding to the above mentioned test jobs.

On Wed, Jul 10, 2024 at 8:45 PM maisebastian @.***> wrote:

Dear Mahesh, thanks for the detailed report. It seems that in your version of OpenMolcas, the format in RASSI is minimally different, the dipole moments are given as PROPERTY: MLTPL 1 COMPONENT: 1 whereas in our installations it is PROPERTY: MLTPL 1 COMPONENT:* 1. The missing little asterisk in your output stopped the interface from finding the data.

Can you please download https://github.com/sharc-md/sharc/blob/sharc3preview/bin/SHARC_MOLCAS.py and try whether it fixes your problem?

Best, Sebastian

— Reply to this email directly, view it on GitHub https://github.com/sharc-md/sharc/issues/131#issuecomment-2220781158, or unsubscribe https://github.com/notifications/unsubscribe-auth/BJY22RL7IKDSRZI2Z6NR5SLZLVF2PAVCNFSM6AAAAABKU5IGN2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMRQG44DCMJVHA . You are receiving this because you authored the thread.Message ID: @.***>

-- Regards,

Mahesh Gudem

Assistant Professor Department of Chemistry Indian Institute of Technology (IIT), Dharwad Karnataka 580011

maisebastian commented 1 month ago

Dear Mahesh, I seem to not see your attachment.

But what you write sounds like these might be different errors. So at least the "Found correct RASSI run, but too few matrix elements!" error seems to have been fixed successfully :)

Best, Sebastian

maheshgiitdh commented 1 month ago

Yes, recent errors might be different to the earlier one. Here, I have attached the image and zip file. RUNNING_TESTS.zip sharc_test_job_summary

maisebastian commented 1 month ago

Thanks for the files.

The error for the two tCSDM tests seems to be because there is no run.sh file in these two tests. I already created these files and will push them to the main branch in a few minutes.

For MOLCAS_dyson, I think that the problem is with the wfoverlap.x installation. If you want to compute Dyson norms in SHARC 3, then you need to recompile wfoverlaps.x linked to some libraries of COLUMBUS (in particular, the COLUMBUS-MOLCAS interface). This compilation is extremely complicated, so I just recommend you that you ignore this test, unless you need Dyson norms for your project.

In the next major release, this dependency will be removed, as MOLCAS can compute Dyson norms natively now.