Polarizability Error: "AO Mux" has an unrecognized shape

Although I have successfully run this command before, I am now running into problems:

psi4.properties("b3lyp-d3bj/aug-cc-pVTZ", properties=["DIPOLE_POLARIZABILITIES", "MULLIKEN_CHARGES", "DIPOLE"])

Here’s an example input where it DOES works for me (for water):

memory 10240 mb

molecule mol {
0 1
O          0.44260       -0.88720       -0.29490
H          0.73940       -0.26640       -0.96690
H          0.79240       -0.54630        0.53350

no_reorient
no_com
}


set scf_type df
set basis aug-cc-pVTZ
set reference rks
set_num_threads(32)

psi4.properties("b3lyp-d3bj/aug-cc-pVTZ", properties=["DIPOLE_POLARIZABILITIES", "MULLIKEN_CHARGES", "DIPOLE"])
mol.print_out()
print_variables()

Here’s an example input where it DOES NOT works for me (for this well-known molecule: https://pubchem.ncbi.nlm.nih.gov/compound/12695449)

memory 102400 mb

molecule mol {
0 1
C 0.00000000 0.00000000 0.00000000
O 1.26010045 -0.67269842 -0.00046169
C 2.33120144 0.16040500 0.00123217
C 3.60030165 -0.61719424 0.00061854
C 4.80980163 0.09100566 0.00218593
C 6.04220157 -0.58379452 0.00170931
C 6.05890153 -1.99539468 -0.00037981
C 4.84350168 -2.69719415 -0.00194021
C 3.62350173 -2.01559399 -0.00144821
H 2.70170162 -2.59169436 -0.00269676
H 4.83690169 -3.78539412 -0.00355902
C 7.29110130 -2.66919546 -0.00085504
H 7.30500122 -3.75769574 -0.00246555
C 8.50770111 -1.96759615 0.00070558
C 8.49100115 -0.55589598 0.00279485
C 7.25860142 0.11770495 0.00326969
H 7.24720141 1.20650493 0.00488172
C 9.70760096 0.14570333 0.00435547
H 9.71540103 1.23340360 0.00597404
C 10.9247008 -0.53679727 0.00386089
H 11.8587006 0.01880225 0.00508420
C 10.9415008 -1.92819718 0.00180180
H 11.8887008 -2.46109735 0.00141430
C 9.74100101 -2.63989649 0.00023305
H 9.77560101 -3.72709649 -0.00136670
H 4.78920141 1.18010491 0.00379448
O 2.28650191 1.38050665 0.00302498
H -0.78830091 -0.75790318 -0.00146161
H -0.10569879 0.61440421 -0.90233546
H -0.10646900 0.61172168 0.90406581

no_reorient
no_com
}


set scf_type df
set basis aug-cc-pVTZ
set reference rks
set_num_threads(64)

psi4.properties("b3lyp-d3bj/aug-cc-pVTZ", properties=["DIPOLE_POLARIZABILITIES", "MULLIKEN_CHARGES", "DIPOLE"])
mol.print_out()
print_variables()

The punchline seems to be:

AO Mux" has an unrecognized shape

but I honestly don’t know. The full fail output is as follows (thanks in advance for any help!):

   ==> Requested Responses <==

    Dipole polarizabilities

Traceback (most recent call last):
  File "/data/Apps/anaconda3/bin/psi4", line 287, in <module>
    exec(content)
  File "<string>", line 60, in <module>
  File "/data/Apps/anaconda3/lib//python3.7/site-packages/psi4/driver/driver.py", line 816, in properties
    return driver_cbs._cbs_gufunc(properties, lowername, ptype='properties', **kwargs)
  File "/data/Apps/anaconda3/lib//python3.7/site-packages/psi4/driver/driver_cbs.py", line 1949, in _cbs_gufunc
    ptype_value, wfn = func(method_name, return_wfn=True, molecule=molecule, **kwargs)
  File "/data/Apps/anaconda3/lib//python3.7/site-packages/psi4/driver/driver.py", line 826, in properties
    wfn = procedures['properties'][lowername](lowername, **kwargs)
  File "/data/Apps/anaconda3/lib//python3.7/site-packages/psi4/driver/procrouting/proc.py", line 2485, in run_scf_property
    print_lvl = (core.get_global_option("PRINT") + 1))
  File "/data/Apps/anaconda3/lib//python3.7/site-packages/psi4/driver/procrouting/response/scf_response.py", line 171, in cpscf_linear_response
    vector_names[i], nbf, nbf, ndocc, nvirt))

psi4.driver.p4util.exceptions.ValidationError: ERROR: "AO Mux" has an unrecognized shape. Must be either (1103, 1103) or (62, 1041)


Printing out the relevant lines from the Psithon --> Python processed input file:
    core.IO.set_default_namespace("mol")
    core.set_global_option("SCF_TYPE", "df")
    core.set_global_option("BASIS", "aug-cc-pVTZ")
    core.set_global_option("REFERENCE", "rks")
    set_num_threads(64)
--> psi4.properties("b3lyp-d3bj/aug-cc-pVTZ", properties=["DIPOLE_POLARIZABILITIES", "MULLIKEN_CHARGES", "DIPOLE"])
    mol.print_out()
    print_variables()

    Psi4 stopped on: Friday, 28 August 2020 04:54PM
    Psi4 wall time for execution: 0:04:46.92

*** Psi4 encountered an error. Buy a developer more coffee!
*** Resources and help at github.com/psi4/psi4.
Traceback (most recent call last):
  File "/data/Apps/anaconda3/bin/psi4", line 287, in <module>
    exec(content)
  File "<string>", line 60, in <module>
  File "/data/Apps/anaconda3/lib//python3.7/site-packages/psi4/driver/driver.py", line 816, in properties
    return driver_cbs._cbs_gufunc(properties, lowername, ptype='properties', **kwargs)
  File "/data/Apps/anaconda3/lib//python3.7/site-packages/psi4/driver/driver_cbs.py", line 1949, in _cbs_gufunc
    ptype_value, wfn = func(method_name, return_wfn=True, molecule=molecule, **kwargs)
  File "/data/Apps/anaconda3/lib//python3.7/site-packages/psi4/driver/driver.py", line 826, in properties
    wfn = procedures['properties'][lowername](lowername, **kwargs)
  File "/data/Apps/anaconda3/lib//python3.7/site-packages/psi4/driver/procrouting/proc.py", line 2485, in run_scf_property
    print_lvl = (core.get_global_option("PRINT") + 1))
  File "/data/Apps/anaconda3/lib//python3.7/site-packages/psi4/driver/procrouting/response/scf_response.py", line 171, in cpscf_linear_response
    vector_names[i], nbf, nbf, ndocc, nvirt))

psi4.driver.p4util.exceptions.ValidationError: ERROR: "AO Mux" has an unrecognized shape. Must be either (1103, 1103) or (62, 1041)

Printing out the relevant lines from the Psithon --> Python processed input file:
    core.IO.set_default_namespace("mol")
    core.set_global_option("SCF_TYPE", "df")
    core.set_global_option("BASIS", "aug-cc-pVTZ")
    core.set_global_option("REFERENCE", "rks")
    set_num_threads(64)
--> psi4.properties("b3lyp-d3bj/aug-cc-pVTZ", properties=["DIPOLE_POLARIZABILITIES", "MULLIKEN_CHARGES", "DIPOLE"])
    mol.print_out()
    print_variables()

Hi,

I cannot reproduce the error. Neither with v1.3.2 nor the dev version.
Did you specify a scratch directory? If not try it out (export PSI_SCRATCH=/path/ or psi4 -s /path/ <options>)

this is my bad. I posted the input file for H2O, which DOES work (I should have clarified).

The following is an example of a molecule that DOES NOT work for me:

memory 102400 mb

molecule mol {
0 1
C 0.00000000 0.00000000 0.00000000
O 1.26010045 -0.67269842 -0.00046169
C 2.33120144 0.16040500 0.00123217
C 3.60030165 -0.61719424 0.00061854
C 4.80980163 0.09100566 0.00218593
C 6.04220157 -0.58379452 0.00170931
C 6.05890153 -1.99539468 -0.00037981
C 4.84350168 -2.69719415 -0.00194021
C 3.62350173 -2.01559399 -0.00144821
H 2.70170162 -2.59169436 -0.00269676
H 4.83690169 -3.78539412 -0.00355902
C 7.29110130 -2.66919546 -0.00085504
H 7.30500122 -3.75769574 -0.00246555
C 8.50770111 -1.96759615 0.00070558
C 8.49100115 -0.55589598 0.00279485
C 7.25860142 0.11770495 0.00326969
H 7.24720141 1.20650493 0.00488172
C 9.70760096 0.14570333 0.00435547
H 9.71540103 1.23340360 0.00597404
C 10.9247008 -0.53679727 0.00386089
H 11.8587006 0.01880225 0.00508420
C 10.9415008 -1.92819718 0.00180180
H 11.8887008 -2.46109735 0.00141430
C 9.74100101 -2.63989649 0.00023305
H 9.77560101 -3.72709649 -0.00136670
H 4.78920141 1.18010491 0.00379448
O 2.28650191 1.38050665 0.00302498
H -0.78830091 -0.75790318 -0.00146161
H -0.10569879 0.61440421 -0.90233546
H -0.10646900 0.61172168 0.90406581

no_reorient
no_com
}

set scf_type df
set basis aug-cc-pVTZ
set reference rks
set_num_threads(64)

psi4.properties(“b3lyp-d3bj/aug-cc-pVTZ”, properties=[“DIPOLE_POLARIZABILITIES”, “MULLIKEN_CHARGES”, “DIPOLE”])
mol.print_out()
print_variables()

I can reproduce it. There is a bug that occurs when linear dependencies in the basis are removed.

To continue with your calculations, either use a less diffuse basis set (e.g. jun-cc-pVTZ) or increase the overlap tolerance to not remove any MOs and the risk of SCF troubles:
set s_tolerance 1e-9

1 Like

A bug fix has been submitted and will soon be available in the development version: https://github.com/psi4/psi4/pull/1999

1 Like

thanks very much! The ‘set s_tolerance 1e-9’ did the trick.

Please mark it solved. Thanks.