Losing symmetry on first step of geometry optimization

#1

I am trying to optimize the Cs symmetry geometry of the benzoic acid and triethylamine complex but I get the following error. The initial geometry is fully optimized using Gaussian (wB97X-D/aDZ). The input file is also appended.

====== last few lines of the output file ==============================

        ERROR: Symmetry operation 1 did not map atom 1 to another atom:
          attempted to find atom at    6.296341 2.199168 -0.689321
          atom_map() throwing PsiException
        Unable to symmetrize geometry.
        Increasing symmetry tolerance to     0.0750
        ERROR: Symmetry operation 1 did not map atom 1 to another atom:
          attempted to find atom at    6.296341 2.199168 -0.689321
          atom_map() throwing PsiException
        Unable to symmetrize geometry.
        Increasing symmetry tolerance to     0.1125
        ERROR: Symmetry operation 1 did not map atom 1 to another atom:
          attempted to find atom at    6.296341 2.199168 -0.689321
          atom_map() throwing PsiException
        Unable to symmetrize geometry.
        Increasing symmetry tolerance to     0.1688
        ERROR: Symmetry operation 1 did not map atom 1 to another atom:
          attempted to find atom at    6.296341 2.199168 -0.689321
          atom_map() throwing PsiException
        Unable to symmetrize geometry.
        Increasing symmetry tolerance to     0.2531
        ERROR: Symmetry operation 1 did not map atom 1 to another atom:
          attempted to find atom at    6.296341 2.199168 -0.689321
          atom_map() throwing PsiException
        Unable to symmetrize geometry.
        Increasing symmetry tolerance to     0.3797
        ERROR: Symmetry operation 1 did not map atom 1 to another atom:
          attempted to find atom at    6.296341 2.199168 -0.689321
          atom_map() throwing PsiException
        Unable to symmetrize geometry.
        Increasing symmetry tolerance to     0.5695
        ERROR: Symmetry operation 1 did not map atom 1 to another atom:
          attempted to find atom at    6.296341 2.199168 -0.689321
          atom_map() throwing PsiException
        Unable to symmetrize geometry.
        Increasing symmetry tolerance to     0.8543
        ERROR: Symmetry operation 1 did not map atom 1 to another atom:
          attempted to find atom at    6.296341 2.199168 -0.689321
          atom_map() throwing PsiException
        Unable to symmetrize geometry.
        Increasing symmetry tolerance to     1.2814
        ERROR: Symmetry operation 1 did not map atom 1 to another atom:
          attempted to find atom at    6.296341 2.199168 -0.689321
          atom_map() throwing PsiException
        Unable to symmetrize geometry.
        Increasing symmetry tolerance to     1.9222
        ERROR: Symmetry operation 1 did not map atom 2 to another atom:
          attempted to find atom at    5.729294 2.867505 -2.658083
          atom_map() throwing PsiException
        Unable to symmetrize geometry.

  **** Optimization has failed! (in 1 steps) ****
                        --------------------------
                         OPTKING Finished Execution
                        --------------------------
        Removing binary optimization data file.
        Cleaning optimization helper files.

PsiException: Could not converge geometry optimization in 0 iterations.


Traceback (most recent call last):
  File "/mnt/home/kakeshpo/anaconda3/bin/psi4", line 269, in <module>
    exec(content)
  File "<string>", line 66, in <module>
  File "/mnt/home/kakeshpo/anaconda3/lib//python3.6/site-packages/psi4/driver/driver.py", line 1220, in optimize
    raise OptimizationConvergenceError("""geometry optimization""", n - 1, wfn)

OptimizationConvergenceError: Could not converge geometry optimization in 0 iterations.

    Psi4 stopped on: Friday, 18 January 2019 01:55AM
    Psi4 wall time for execution: 0:08:59.58

*** Psi4 encountered an error. Buy a developer more coffee!
*** Resources and help at github.com/psi4/psi4.

====== end of output file ============================

====== input file ==================================

molecule {
0 1
 C                  3.21688113    1.15174036    0.00000000
 H                  2.92869003    1.74850938    0.87371200
 H                  2.92869003    1.74850938   -0.87371200
 N                  2.38500072   -0.05936026    0.00000000
 C                  4.72833918    0.92738468    0.00000000
 H                  5.05878731    0.37491737   -0.88927900
 H                  5.05878731    0.37491737    0.88927900
 H                  5.24276530    1.89689042    0.00000000
 C                  2.53800152   -0.89618441   -1.20087800
 H                  1.79768853   -1.69943379   -1.10018700
 H                  3.53587423   -1.37075354   -1.22212200
 C                  2.53800152   -0.89618441    1.20087800
 H                  1.79768853   -1.69943379    1.10018700
 H                  3.53587423   -1.37075354    1.22212200
 C                  2.28578921   -0.14932674    2.50359900
 H                  2.23254080   -0.87093862    3.32808900
 H                  1.33177849    0.39281482    2.46491100
 H                  3.08478055    0.56454553    2.74249300
 C                  2.28578921   -0.14932674   -2.50359900
 H                  1.33177849    0.39281482   -2.46491100
 H                  2.23254080   -0.87093862   -3.32808900
 H                  3.08478055    0.56454553   -2.74249300
 O                 -0.15805335    0.68773704    0.00000000
 H                  0.82616689    0.38854021    0.00000000
 C                 -0.94154401   -0.37928711    0.00000000
 O                 -0.52146693   -1.52491092    0.00000000
 C                 -2.40325702   -0.05775038    0.00000000
 C                 -5.14373328    0.46823462    0.00000000
 C                 -3.31998328   -1.11310970    0.00000000
 C                 -2.86282721    1.26273104    0.00000000
 C                 -4.23138203    1.52350341    0.00000000
 C                 -4.68684583   -0.85059427    0.00000000
 H                 -2.94184553   -2.13429164    0.00000000
 H                 -2.14358229    2.07932838    0.00000000
 H                 -4.58802597    2.55348371    0.00000000
 H                 -5.39944287   -1.67535684    0.00000000
 H                 -6.21453295    0.67401418    0.00000000
}


memory 60 GB

set basis aug-cc-pvdz
set scf_type df
set freeze_core True

optimize('mp2')
freq('mp2')

===== end of input file =============================

0 Likes

#2

when I include “symmetry Cs” in molecule specifications, then i get this error:

====last a few lines of output ====================

Fatal Error: User specified point group (Cs(X)) is not a subgroup of the highest detected point group (C1). If this is because the symmetry increased, try to start the calculation again from the last geometry, after checking any symmetry-dependent input, such as DOCC.
Error occurred in file: /scratch/psilocaluser/conda-builds/psi4-multiout_1532493090788/work/psi4/src/psi4/libmints/molecule.cc on line: 2625
The most recent 5 function calls were:

psi::Molecule::find_point_group(double) const
psi::Molecule::update_geometry()

Traceback (most recent call last):
  File "/mnt/home/kakeshpo/anaconda3/bin/psi4", line 269, in <module>
    exec(content)
  File "<string>", line 67, in <module>
  File "/mnt/home/kakeshpo/anaconda3/lib//python3.6/site-packages/psi4/driver/driver.py", line 1167, in optimize
    moleculeclone.update_geometry()

RuntimeError:
Fatal Error: User specified point group (Cs(X)) is not a subgroup of the highest detected point group (C1). If this is because the symmetry increased, try to start the calculation again from the last geometry, after checking any symmetry-dependent input, such as DOCC.
Error occurred in file: /scratch/psilocaluser/conda-builds/psi4-multiout_1532493090788/work/psi4/src/psi4/libmints/molecule.cc on line: 2625
The most recent 5 function calls were:

psi::Molecule::find_point_group(double) const
psi::Molecule::update_geometry()


    Psi4 stopped on: Friday, 18 January 2019 02:34AM
    Psi4 wall time for execution: 0:01:34.06

*** Psi4 encountered an error. Buy a developer more coffee!
*** Resources and help at github.com/psi4/psi4.

===== end of output file =============================

0 Likes

#3

I made a couple edits to your posts. I’ve enclosed all your sample code in triple backticks (`) so that the forum formats it correctly, and I’ve changed the topic title to be more specific.

I can reproduce your error on a recent version of Psi. It looks like something is wrong with the gradient. Here’s the HF gradient I get:

                     0.0000214140       -0.0000305614        0.0000000000
                     0.0000025047        0.0000069054        0.0000026313
                     0.0000025047        0.0000069054       -0.0000026313
                    -0.0000137691        0.0000724762        0.0000000000
                     0.0000148350        0.0000037577        0.0000000000
                     0.0000027477       -0.0000004823       -0.0000034740
                     0.0000027477       -0.0000004823        0.0000034740
                    -0.0000029663        0.0000042586        0.0000000000
                    -0.0000095301       -0.0000075367       -0.0000017055
                    -0.0000046504       -0.0000041348        0.0000021955
                    -0.0000016483       -0.0000010402        0.0000003236
                    -0.0000095301       -0.0000075367        0.0000017055
                    -0.0000046504       -0.0000041348       -0.0000021955
                    -0.0000016483       -0.0000010402       -0.0000003236
                     0.0000001687        0.0000032351        0.0000154260
                     0.0000000771       -0.0000033085        0.0000051330
                    -0.0000030464        0.0000000940       -0.0000027541
                     0.0000049561        0.0000057538        0.0000034163
                     0.0000001687        0.0000032351       -0.0000154260
                    -0.0000030464        0.0000000940        0.0000027541
                     0.0000000771       -0.0000033085       -0.0000051330
                     0.0000049561        0.0000057538       -0.0000034163
                    -0.0000526002       -0.0000491134        0.0000000000
                    -0.0000169703       -0.0000464658        0.0000000000
                     0.0000894944        0.0001047309        0.0000000000
                    -0.0000139172       -0.0000660825        0.0000000000
                    -0.0000190128        0.0000062307        0.0000000000
                    -0.0000077694       -0.0000009124        0.0000000000
                     0.0000368255       -0.0000018760        0.0000000000
                     0.0000044643       -0.0000094501        0.0000000000
                    -0.0000046154        0.0000080775        0.0000000000
                    -0.0000094790        0.0000048102        0.0000000000
                    -0.0000030777       -0.0000081173        0.0000000000
                     0.0000047611        0.0000082287        0.0000000000
                    -0.0000002769        0.0000078923        0.0000000000
                    -0.0000040740       -0.0000074846        0.0000000000
                    -0.0000064242        0.0000006292        0.0000000000

Note that there are a lot of zeros and positive/negative pairs due to the Cs symmetry. But look at the MP2 gradient.

                    -0.0013181670       -0.0018271400        0.0000002361
                     0.0009011354       -0.0013054532       -0.0027104992
                     0.0009011237       -0.0013054667        0.0027104637
                     0.0084081544       -0.0028925304        0.0000000001
                    -0.0010543251        0.0012503834        0.0000000063
                    -0.0006421599        0.0011696981        0.0018253307
                    -0.0006421578        0.0011697002       -0.0018253281
                    -0.0012125068       -0.0023019874       -0.0000000082
                    -0.0016803862        0.0021654515        0.0007771133
                     0.0020188548        0.0023175515       -0.0008096447
                    -0.0026838052        0.0012891903       -0.0001062773
                    -0.0016802763        0.0021661454       -0.0007772097
                     0.0020187584        0.0023173420        0.0008095331
                    -0.0026838317        0.0012890861        0.0001061677
                    -0.0000475013       -0.0021122278       -0.0007453141
                     0.0000651903        0.0017584727       -0.0018191311
                     0.0021150089       -0.0011541165        0.0003155535
                    -0.0014345486       -0.0013660345       -0.0003146311
                    -0.0000475521       -0.0021123561        0.0007454293
                     0.0021150439       -0.0011540327       -0.0003156145
                     0.0000651534        0.0017583782        0.0018191773
                    -0.0014345213       -0.0013660218        0.0003146042
                    -0.0016331680       -0.0177542470        0.0000002390
                    -0.0061401950        0.0036687120       -0.0000001893
                     0.0102779369       -0.0068153971        0.0000000100
                    -0.0075592108        0.0227241740       -0.0000000891
                    -0.0072489349        0.0007529733        0.0000000849
                     0.0110601784       -0.0020605684       -0.0000000026
                    -0.0040874598        0.0098453268       -0.0000000146
                    -0.0073783354       -0.0080263850        0.0000000100
                     0.0026985355       -0.0082104306       -0.0000000028
                     0.0056731377        0.0065637136        0.0000000050
                    -0.0011826084        0.0027309606       -0.0000000037
                    -0.0021329197       -0.0021112118       -0.0000000038
                     0.0009792989       -0.0026397473       -0.0000000006
                     0.0018726309        0.0021021766        0.0000000002
                     0.0027544301       -0.0005240819        0.0000000003

Many components that should be zero or positive/negative pairs are not. The deviation is slight, but it’s enough to cost you symmetry. It looks to me like this gradient was never adapted to point group symmetry, but somebody more familiar with the code should check me on this.

0 Likes

#4

Confirmed, and fixed in this pull request. Thanks for the report!

0 Likes

#5

Thank you. I will consider these points for future questions.

0 Likes

#6

That is great. How do I add this bug fix to my Psi4? I am using the conda version. Would it be included if I update the package?

0 Likes

#7

Not yet. We need to accept the fix (which should be soon) and we need to update the nightly channel of conda. Then on your end, it’s a simple matter of switching to the nightly version if necessary and updating the package.

We’ll post here when both of those things on the dev end are finished.

0 Likes

#8

It’s taken a while, but we have the developer version of Psi4 on conda!

Follow the instructions for Linux -> Conda -> Nightly here. If those work, mark a post here as closing the issue, so we know the problem is solved. If not, let us know what the new problem is.

0 Likes

#9

Thank you. I am ready to use the following command to install the nightly version.

 conda install psi4 psi4-rt python=3.6 -c psi4/label/dev

but before doing that I have a question: Does installing the nightly version affect my stable version? and can I use them in parallel on the same system?

0 Likes

#10

That command would replace the stable version with the nightly version. If you want access to both, change that command to

conda create -n newenv psi4 psi4-rt python=3.6 -c psi4/label/dev

newenv is just a name. You can change to whatever you want. The important part is that the command will now install the developer version of Psi to a new conda environment, separate from your other installs. When you want to use stable Psi, just use it as normal. When you want to use developer Psi, activate the conda environment newenv and run Psi as normal to. When you want to use normal Psi again, deactivate the environment. (See here.)

0 Likes

#11

I was able to install the nightly version on a new environment (newenv), but when I tried to execute a job, I got the following error:

Traceback (most recent call last):
  File "/mnt/home/kakeshpo/anaconda3/envs/newenv/bin/psi4", line 177, in <module>
    import psi4
  File "/mnt/home/kakeshpo/anaconda3/envs/newenv/lib//python3.6/site-packages/psi4/__init__.py", line 82, in <module>
    from .driver import endorsed_plugins
  File "/mnt/home/kakeshpo/anaconda3/envs/newenv/lib//python3.6/site-packages/psi4/driver/endorsed_plugins.py", line 33, in <module>
    import v2rdm_casscf
  File "/mnt/home/kakeshpo/anaconda3/envs/newenv/lib//python3.6/site-packages/v2rdm_casscf/__init__.py", line 45, in <module>
    psi4.core.plugin_load(sofile)
RuntimeError:
Fatal Error: load_plugin: Cannot open library: /mnt/home/kakeshpo/anaconda3/envs/newenv/lib/python3.6/site-packages/v2rdm_casscf/v2rdm_casscf.so: undefined symbol: _ZN3psi6Vector8subtractERKSt10shared_ptrIS0_E
Error occurred in file: /scratch/psilocaluser/conda-builds/psi4-multiout_1548756333092/work/psi4/src/psi4/libplugin/load_plugin.cc on line: 52
The most recent 5 function calls were:


forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image              PC                Routine            Line        Source
libgdma.so         00002AD2BD70A0DC  for__signal_handl     Unknown  Unknown
libpthread-2.17.s  00002AD2B1DB45E0  Unknown               Unknown  Unknown
ld-2.17.so         00002AD2B1B962F1  Unknown               Unknown  Unknown
ld-2.17.so         00002AD2B1B902D4  Unknown               Unknown  Unknown
libdl-2.17.so      00002AD2B23855BD  Unknown               Unknown  Unknown
libdl-2.17.so      00002AD2B238509F  dlclose               Unknown  Unknown
core.cpython-36m-  00002AD2B9ACBF57  Unknown               Unknown  Unknown
core.cpython-36m-  00002AD2B9CD1A52  Unknown               Unknown  Unknown
core.cpython-36m-  00002AD2B9CF8F16  Unknown               Unknown  Unknown
core.cpython-36m-  00002AD2B97A0F46  Unknown               Unknown  Unknown
python3.6          0000557FD9D6A6A6  PyCFunction_Call      Unknown  Unknown
python3.6          0000557FD9D6754E  PyObject_Call         Unknown  Unknown
python3.6          0000557FD9E684B4  Unknown               Unknown  Unknown
python3.6          0000557FD9E6573F  Py_FinalizeEx         Unknown  Unknown
python3.6          0000557FD9E701FE  Py_Main               Unknown  Unknown
python3.6          0000557FD9D3902E  main                  Unknown  Unknown
libc-2.17.so       00002AD2B1FE2C05  __libc_start_main     Unknown  Unknown
python3.6          0000557FD9E19E0E  Unknown               Unknown  Unknown
0 Likes

#12

There are a few fixes, but the simplest one is just conda remove v2rdm_casscf while in newenv.

0 Likes

#13

Thank you! I was able to optimize my geometry successfully. I am running frequencies and so far I have not seen any error.

  --------------------------------------------------------------------------------------------- ~
   Step     Total Energy     Delta E     MAX Force     RMS Force      MAX Disp      RMS Disp    ~
  --------------------------------------------------------------------------------------------- ~
    Convergence Criteria    1.00e-06 *    3.00e-04 *             o    1.20e-03 *             o  ~
  --------------------------------------------------------------------------------------------- ~
      1    -711.23285635   -7.11e+02      2.29e-02      3.21e-03 o    3.83e-02      8.75e-03 o  ~
      2    -711.23491866   -2.06e-03      3.37e-03      4.57e-04 o    2.77e-02      5.29e-03 o  ~
      3    -711.23504772   -1.29e-04      2.56e-03      2.43e-04 o    4.77e-02      7.42e-03 o  ~
      4    -711.23513030   -8.26e-05      1.71e-03      2.01e-04 o    1.94e-02      4.58e-03 o  ~
      5    -711.23515698   -2.67e-05      9.91e-04      1.13e-04 o    2.14e-02      3.61e-03 o  ~
      6    -711.23517191   -1.49e-05      3.08e-04      5.06e-05 o    6.08e-03      1.29e-03 o  ~
      7    -711.23517408   -2.17e-06      2.19e-04 *    2.31e-05 o    1.95e-03      4.35e-04 o  ~
      8    -711.23517463   -5.50e-07 *    1.34e-04 *    1.40e-05 o    2.06e-03      2.99e-04 o  ~
0 Likes

#14

Glad to hear it! Feel free to post again if any more symmetry-loss problems come up. Seeing as this works for you, I’m going to mark this topic as solved.

0 Likes

#15

Thank you for your help! So far the frequencies are running and there is no error. I will know in a couple of days for sure if the frequencies finish successfully :crossed_fingers:.

0 Likes

#16

Thank you.
Had the same bug as Tayeb, the workaround to remove v2rdm_casscf worked and example does work as expected.

0 Likes

Geometry optimization fails due to changing comp. point group
#17

Good!

Incidentally, the stable version won’t need you to remove v2rdm_casscf in the first place.

0 Likes

closed #18

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.

0 Likes