dftbplus
dftbplus copied to clipboard
Published
20 hours ago •
dftbplus
dftbplus
DFTB+ 21.2 with xTB Error: Diagonalization fails in dsyevr
Starting initialization...
--------------------------------------------------------------------------------
OpenMP threads: 2
Chosen random seed: 585458
Mode: LBFGS relaxation
Self consistent charges: Yes
SCC-tolerance: 0.100000E-04
Max. scc iterations: 500
Shell resolved Hubbard: Yes
Spin polarisation: Yes
Initial nr. of up electrons: 878.000000
Initial Nr. of down electrons 878.000000
Periodic boundaries: No
Electronic solver: Relatively robust
Mixer: Broyden mixer
Mixing parameter: 0.200000
Maximal SCC-cycles: 500
Nr. of chrg. vec. in memory: 500
Nr. of moved atoms: 213
Max. nr. of geometry steps: 500
Force tolerance: 0.100000E-01
tblite library version: 0.1.0
-> parametrization: GFN1-xTB
-> repulsion: Yes
-> dispersion: Yes
-> halogen bonding: Yes
-> electrostatics: Yes
-> isotropic: Yes
-> anisotropic: No
-> third-order: Yes
Electronic temperature: 0.950045E-03 H 0.258520E-01 eV
Initial charges: Set automatically (system chrg: 0.000E+00)
Included shells: Fe: d, s, p
O: s, p
H: s, s2
C: s, p
N: s, p
S: s, p, d
Spin coupling constants: Fe 1(d)-1(d): -0.150000E-01
Fe 1(d)-2(s): -0.120000E-01
Fe 1(d)-3(p): -0.300000E-02
Fe 2(s)-1(d): -0.120000E-01
Fe 2(s)-2(s): -0.190000E-01
Fe 2(s)-3(p): -0.100000E-02
Fe 3(p)-1(d): -0.300000E-02
Fe 3(p)-2(s): -0.100000E-02
Fe 3(p)-3(p): -0.170000E-01
O 1(s)-1(s): -0.350000E-01
O 1(s)-2(p): -0.300000E-01
O 2(p)-1(s): -0.300000E-01
O 2(p)-2(p): -0.280000E-01
H 1(s)-1(s): -0.720000E-01
H 1(s)-2(s): -0.670000E-01
H 2(s)-1(s): -0.670000E-01
H 2(s)-2(s): -0.720000E-01
C 1(s)-1(s): -0.310000E-01
C 1(s)-2(p): -0.250000E-01
C 2(p)-1(s): -0.250000E-01
C 2(p)-2(p): -0.230000E-01
N 1(s)-1(s): -0.330000E-01
N 1(s)-2(p): -0.270000E-01
N 2(p)-1(s): -0.270000E-01
N 2(p)-2(p): -0.260000E-01
S 1(s)-1(s): -0.210000E-01
S 1(s)-2(p): -0.170000E-01
S 1(s)-3(d): 0.000000E+00
S 2(p)-1(s): -0.170000E-01
S 2(p)-2(p): -0.160000E-01
S 2(p)-3(d): 0.000000E+00
S 3(d)-1(s): 0.000000E+00
S 3(d)-2(p): 0.000000E+00
S 3(d)-3(d): -0.800000E-01
Extra options:
Mulliken analysis
Force type original
--------------------------------------------------------------------------------
*** Geometry step: 0
iSCC Total electronic Diff electronic SCC error
1 0.00000000E+00 0.00000000E+00 0.30430713E+16
2 -0.82820354E+45 -0.82820354E+45 0.24436506+110
ERROR!
-> Failure in DSYEVR 1
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_OVERFLOW_FLAG IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
ERROR STOP
Error termination. Backtrace:
#0 0x14c61d558171 in ???
#1 0x14c61d558d19 in ???
#2 0x14c61d559e8e in ???
#3 0x409c7e in __dftbp_common_globalenv_MOD_abortprogram
at /root/dftbplus-21.2/src/dftbp/common/globalenv.F90:195
#4 0x5ced81 in __dftbp_math_eigensolver_MOD_dble_dsygvr
at /root/dftbplus-21.2/src/dftbp/math/eigensolver.F90:1211
#5 0x8b9d55 in __dftbp_dftbplus_eigenvects_MOD_diagdenserealmtx
at /root/dftbplus-21.2/src/dftbp/dftbplus/eigenvects.F90:86
#6 0x479546 in buildanddiagdenserealham
at /root/dftbplus-21.2/src/dftbp/dftbplus/main.F90:2757
#7 0x4a0f79 in getdensityfromdensediag
at /root/dftbplus-21.2/src/dftbp/dftbplus/main.F90:2591
#8 0x4a862f in getdensity
at /root/dftbplus-21.2/src/dftbp/dftbplus/main.F90:2406
#9 0x4b18b7 in __dftbp_dftbplus_main_MOD_processgeometry
at /root/dftbplus-21.2/src/dftbp/dftbplus/main.F90:963
#10 0x4bb09a in __dftbp_dftbplus_main_MOD_rundftbplus
at /root/dftbplus-21.2/src/dftbp/dftbplus/main.F90:224
#11 0x4080db in dftbplus
at /root/dftbplus-21.2/app/dftb+/dftbplus.F90:34
#12 0x4048cc in main
at /root/dftbplus-21.2/app/dftb+/dftbplus.F90:11
Thanks for reporting! Your system is inedeed very difficult to get converged. By setting a considerable higher electronic temperature, it seems to be possible, though.
However, I don't get the numbers like you do! Using the official non-mpi conda binary, I get much more reasonable energies, and no diagonalization error for the input you've attached. So I am wondering, which binary you are using, and whether the diagonalizer of the binary is OK. Could you please try the conda binary for the same system and see, whether you get a diagonalization error?
One more comment. Your system is highly unphysical. Are you sure, you did not forget the periodic boundary conditions?
Thanks for your volunrable reply！I used Intel(R) Parallel Studio XE 2020 to compile DFTB+ 21.2 and tblite in my Vmware workstation. The version of my system is Centos 8.
@ruauar Did anything happen here? Did you try the official conda-distributed binary? As I wrote above, I get different numbers than you with the official binary...
I am also getting this error with the official conda build but on a different system. Strangely, it doesn't start the first electronic iteration, so I'm not sure if that's related or not. Raising the electronic temperature did not help.
|===============================================================================
|
| DFTB+ release 21.2
|
| Copyright (C) 2006 - 2021 DFTB+ developers group
|
|===============================================================================
|
| When publishing results obtained with DFTB+, please cite the following
| reference:
|
| * DFTB+, a software package for efficient approximate density functional
| theory based atomistic simulations, J. Chem. Phys. 152, 124101 (2020).
| [doi: 10.1063/1.5143190]
|
| You should also cite additional publications crediting the parametrization
| data you use. Please consult the documentation of the SK-files for the
| references.
|
|===============================================================================
Reading input file 'dftb_in.hsd'
Parser version: 10
--------------------------------------------------------------------------------
Processed input in HSD format written to 'dftb_pin.hsd'
Starting initialization...
--------------------------------------------------------------------------------
OpenMP threads: 16
Chosen random seed: 1180113
Mode: Static calculation
Self consistent charges: Yes
SCC-tolerance: 0.100000E-04
Max. scc iterations: 100
Shell resolved Hubbard: Yes
Spin polarisation: No
Nr. of up electrons: 776.000000
Nr. of down electrons: 776.000000
Periodic boundaries: Yes
Electronic solver: Relatively robust
Mixer: Broyden mixer
Mixing parameter: 0.200000
Maximal SCC-cycles: 100
Nr. of chrg. vec. in memory: 100
tblite library version: 0.1.0
-> parametrization: GFN2-xTB
-> repulsion: Yes
-> dispersion: Yes
-> halogen bonding: No
-> electrostatics: Yes
-> isotropic: Yes
-> anisotropic: Yes
-> third-order: Yes
Electronic temperature: 0.950045E-03 H 0.258520E-01 eV
Initial charges: Set automatically (system chrg: 0.000E+00)
Included shells: Zn: s, p
O: s, p
C: s, p
H: s
K-points and weights: 1: 0.000000 0.000000 0.000000 1.000000
K-points in absolute space: 1: 0.000000 0.000000 0.000000
Extra options:
Mulliken analysis
Force type original
--------------------------------------------------------------------------------
*** Geometry step: 0
iSCC Total electronic Diff electronic SCC error
ERROR!
-> Failure in DSYEVR 1
dftb_in.hsd.txt geo_end.gen.txt
I am just runningDFTB+ with your input files on my laptop (serial DFTB+ 21.2 from conda). For me, it does not stop, but finishes the SCC cycle and optimizes the geometry as supposed... So it is somewhat hard to guess, what happens.
Thanks for checking. That's quite strange. I am running this on Ubuntu via Windows Subsystem for Linux but same conda build (Python 3.8) you mention. I will try exploring this in more detail now that I know at least someone has gotten the calculation to start correctly.
For conda related errors it is incredible useful to have the environment information (conda list and conda info, best put them in a <details></details> block in a comment). Errors regarding the solver can be manifold, usually it is related to OpenBLAS / MKL discrepancies. You can always switch the backend by installing 'libblas=*=*mkl*' into your environment.
Thanks! Here are the conda details just for context. I'll dig into this more shortly. (I tried putting it in a details block but couldn't figure out the formatting so it was rendered well).
conda info
active environment : py38
active env location : /home/arosen/anaconda3/envs/py38
shell level : 2
user config file : /home/arosen/.condarc
populated config files : /home/arosen/.condarc
conda version : 4.10.3
conda-build version : 3.21.5
python version : 3.8.12.final.0
virtual packages : __linux=5.10.102.1=0
__glibc=2.31=0
__unix=0=0
__archspec=1=x86_64
base environment : /home/arosen/anaconda3 (writable)
conda av data dir : /home/arosen/anaconda3/etc/conda
conda av metadata url : None
channel URLs : https://conda.anaconda.org/conda-forge/linux-64
https://conda.anaconda.org/conda-forge/noarch
https://repo.anaconda.com/pkgs/main/linux-64
https://repo.anaconda.com/pkgs/main/noarch
https://repo.anaconda.com/pkgs/r/linux-64
https://repo.anaconda.com/pkgs/r/noarch
package cache : /home/arosen/anaconda3/pkgs
/home/arosen/.conda/pkgs
envs directories : /home/arosen/anaconda3/envs
/home/arosen/.conda/envs
platform : linux-64
user-agent : conda/4.10.3 requests/2.26.0 CPython/3.8.12 Linux/5.10.102.1-microsoft-standard-WSL2 ubuntu/20.04.2 glibc/2.31
UID:GID : 1000:1000
netrc file : None
offline mode : False
conda list
# packages in environment at /home/arosen/anaconda3/envs/py38:
#
# Name Version Build Channel
_libgcc_mutex 0.1 conda_forge conda-forge
_openmp_mutex 4.5 1_gnu conda-forge
aioitertools 0.10.0 pypi_0 pypi
ambit 0.5.1 hbe8a562_1 psi4
anyio 3.5.0 pypi_0 pypi
arpack 3.7.0 hdefa2d7_2 conda-forge
ase 3.23.0b1 pypi_0 pypi
asttokens 2.0.5 pypi_0 pypi
atomate2 0.0.6 pypi_0 pypi
attrs 21.4.0 pyhd8ed1ab_0 conda-forge
backcall 0.2.0 pypi_0 pypi
bcrypt 3.2.0 pypi_0 pypi
blas 1.0 mkl
boto3 1.21.42 pypi_0 pypi
botocore 1.24.42 pypi_0 pypi
bzip2 1.0.8 h7f98852_4 conda-forge
c-ares 1.18.1 h7f98852_0 conda-forge
ca-certificates 2021.10.8 ha878542_0 conda-forge
cclib 1.7.1 pypi_0 pypi
certifi 2021.10.8 pypi_0 pypi
cffi 1.15.0 py38h3931269_0 conda-forge
charset-normalizer 2.0.12 pypi_0 pypi
chemps2 1.8.10 hbe8a562_0 psi4
chimes-calculator 1.0.0 h0c14de7_1 conda-forge
click 8.1.2 pypi_0 pypi
cryptography 36.0.2 pypi_0 pypi
custodian 2022.1.17 pypi_0 pypi
cycler 0.11.0 pypi_0 pypi
decorator 4.4.2 py_0 conda-forge
dftbplus 21.2 nompi_h8767184_102 conda-forge
dftd3 3.2.1 h84218bc_2 psi4
dftd4 3.3.0 hf06ca72_3 conda-forge
dkh 1.2 h173d85e_2 psi4
dnspython 2.2.1 pypi_0 pypi
executing 0.8.3 pypi_0 pypi
fastapi 0.75.2 pypi_0 pypi
fftw 3.3.10 nompi_h77c792f_102 conda-forge
fireworks 2.0.2 pypi_0 pypi
flask 2.1.1 pypi_0 pypi
flask-paginate 2022.1.8 pypi_0 pypi
fonttools 4.32.0 pypi_0 pypi
future 0.18.2 pypi_0 pypi
gau2grid 2.0.7 hd18ef5c_0 psi4
gawk 5.1.0 h7f98852_0 conda-forge
gcp 2.0.2 he991be0_2 psi4
gdma 2.2.6 h0e1e685_6 psi4
gsl 2.7 he838d99_0 conda-forge
gunicorn 20.1.0 pypi_0 pypi
hdf5 1.10.6 nompi_h6a2412b_1114 conda-forge
idna 3.3 pypi_0 pypi
importlib-metadata 4.11.3 py38h578d9bd_1 conda-forge
importlib_metadata 4.11.3 hd8ed1ab_1 conda-forge
importlib_resources 5.7.1 pyhd8ed1ab_0 conda-forge
iniconfig 1.1.1 pyh9f0ad1d_0 conda-forge
intel-openmp 2021.4.0 h06a4308_3561
ipython 8.2.0 pypi_0 pypi
itsdangerous 2.1.2 pypi_0 pypi
jedi 0.18.1 pypi_0 pypi
jinja2 3.1.1 pypi_0 pypi
jmespath 1.0.0 pypi_0 pypi
jobflow 0.1.8 pypi_0 pypi
jsonschema 4.4.0 pypi_0 pypi
keyutils 1.6.1 h166bdaf_0 conda-forge
kiwisolver 1.4.2 pypi_0 pypi
krb5 1.19.3 h3790be6_0 conda-forge
latexcodec 2.0.1 pypi_0 pypi
ld_impl_linux-64 2.36.1 hea4e1c9_2 conda-forge
libblas 3.9.0 12_linux64_mkl conda-forge
libcblas 3.9.0 12_linux64_mkl conda-forge
libcurl 7.82.0 h7bff187_0 conda-forge
libedit 3.1.20191231 he28a2e2_2 conda-forge
libev 4.33 h516909a_1 conda-forge
libffi 3.4.2 h7f98852_5 conda-forge
libgcc-ng 11.2.0 h1d223b6_15 conda-forge
libgfortran-ng 11.2.0 h69a702a_15 conda-forge
libgfortran5 11.2.0 h5c6108e_15 conda-forge
libgomp 11.2.0 h1d223b6_15 conda-forge
libint2 2.6.0 hc9558a2_9 psi4
liblapack 3.9.0 12_linux64_mkl conda-forge
libmbd 0.12.5 nompi_hd3f0df1_101 conda-forge
libnegf 1.0.3 nompi_heef902a_102 conda-forge
libnghttp2 1.47.0 h727a467_0 conda-forge
libnsl 2.0.0 h7f98852_0 conda-forge
libssh2 1.10.0 ha56f1ee_2 conda-forge
libstdcxx-ng 11.2.0 he4da1e4_15 conda-forge
libuuid 2.32.1 h7f98852_1000 conda-forge
libxc 5.1.5 h84b9e52_1 psi4
libzlib 1.2.11 h166bdaf_1014 conda-forge
maggma 0.45.0 pypi_0 pypi
markupsafe 2.1.1 pypi_0 pypi
matplotlib 3.5.1 pypi_0 pypi
matplotlib-inline 0.1.3 pypi_0 pypi
mctc-lib 0.2.5 hc3ea6d6_0 conda-forge
mkl 2021.4.0 h06a4308_640
mkl-service 2.4.0 py38h95df7f1_0 conda-forge
mkl_fft 1.3.1 py38h8666266_1 conda-forge
mkl_random 1.2.2 py38h1abd341_0 conda-forge
mongogrant 0.3.3 pypi_0 pypi
mongomock 4.0.0 pypi_0 pypi
monty 2022.3.12 pypi_0 pypi
mpmath 1.2.1 pypi_0 pypi
msgpack 1.0.3 pypi_0 pypi
ncurses 6.3 h27087fc_1 conda-forge
networkx 2.8 pypi_0 pypi
numpy 1.22.3 pypi_0 pypi
numpy-base 1.21.5 py38hf524024_1
openssl 1.1.1n h166bdaf_0 conda-forge
orjson 3.6.8 pypi_0 pypi
packaging 21.3 pyhd8ed1ab_0 conda-forge
palettable 3.3.0 pypi_0 pypi
pandas 1.4.2 pypi_0 pypi
paramiko 2.10.3 pypi_0 pypi
parso 0.8.3 pypi_0 pypi
pcmsolver 1.2.1.1 py38h6d17ec8_2 psi4
periodictable 1.6.0 pypi_0 pypi
pexpect 4.8.0 pypi_0 pypi
pickleshare 0.7.5 pypi_0 pypi
pillow 9.1.0 pypi_0 pypi
pint 0.17 pyhd8ed1ab_0 psi4
pip 22.0.4 pyhd8ed1ab_0 conda-forge
plotly 5.7.0 pypi_0 pypi
pluggy 1.0.0 py38h578d9bd_3 conda-forge
plumed 2.7.3 nompi_h8de8e23_100 conda-forge
prompt-toolkit 3.0.29 pypi_0 pypi
psi4 1.5+e9f4d6d py38hbf93d9d_0 psi4
psutil 5.9.0 py38h0a891b7_1 conda-forge
ptyprocess 0.7.0 pypi_0 pypi
pure-eval 0.2.2 pypi_0 pypi
py 1.11.0 pyh6c4a22f_0 conda-forge
py-cpuinfo 8.0.0 pyhd8ed1ab_0 conda-forge
pybtex 0.24.0 pypi_0 pypi
pycparser 2.21 pyhd8ed1ab_0 conda-forge
pydantic 1.9.0 py38h0a891b7_1 conda-forge
pydash 5.1.0 pypi_0 pypi
pygments 2.11.2 pypi_0 pypi
pymatgen 2022.4.19 pypi_0 pypi
pymongo 4.1.1 pypi_0 pypi
pynacl 1.5.0 pypi_0 pypi
pyparsing 3.0.8 pyhd8ed1ab_0 conda-forge
pyrsistent 0.18.1 pypi_0 pypi
pytest 7.1.1 py38h578d9bd_1 conda-forge
python 3.8.13 h582c2e5_0_cpython conda-forge
python-dateutil 2.8.2 pypi_0 pypi
python_abi 3.8 2_cp38 conda-forge
pytz 2022.1 pypi_0 pypi
pyyaml 6.0 py38h0a891b7_4 conda-forge
pyzmq 22.3.0 pypi_0 pypi
qcelemental 0.24.0 pyhd8ed1ab_0 psi4
qcengine 0.23.0 pyhd8ed1ab_0 psi4
quacc 0.0.3 dev_0 <develop>
readline 8.1 h46c0cb4_0 conda-forge
requests 2.27.1 pypi_0 pypi
ruamel-yaml 0.17.21 pypi_0 pypi
ruamel-yaml-clib 0.2.6 pypi_0 pypi
s3transfer 0.5.2 pypi_0 pypi
scipy 1.8.0 pypi_0 pypi
sentinels 1.0.0 pypi_0 pypi
sentry-sdk 1.5.10 pypi_0 pypi
setuptools 62.1.0 py38h578d9bd_0 conda-forge
simint 0.7 h642920c_1 psi4
simple-dftd3 0.4.2 hb01ceae_1 conda-forge
six 1.16.0 pyh6c4a22f_0 conda-forge
sniffio 1.2.0 pypi_0 pypi
spglib 1.16.3 pypi_0 pypi
sqlite 3.38.2 h4ff8645_0 conda-forge
sshtunnel 0.4.0 pypi_0 pypi
stack-data 0.2.0 pypi_0 pypi
starlette 0.17.1 pypi_0 pypi
sympy 1.10.1 pypi_0 pypi
tabulate 0.8.9 pypi_0 pypi
tblite 0.0.0 pypi_0 pypi
tenacity 8.0.1 pypi_0 pypi
tk 8.6.12 h27826a3_0 conda-forge
toml-f 0.2.3 hc3ea6d6_0 conda-forge
tomli 2.0.1 pyhd8ed1ab_0 conda-forge
tqdm 4.64.0 pypi_0 pypi
traitlets 5.1.1 pypi_0 pypi
typing-extensions 4.2.0 pypi_0 pypi
typing_extensions 4.1.1 pyha770c72_0 conda-forge
uncertainties 3.1.6 pypi_0 pypi
urllib3 1.26.9 pypi_0 pypi
wcwidth 0.2.5 pypi_0 pypi
werkzeug 2.1.1 pypi_0 pypi
wheel 0.37.1 pyhd8ed1ab_0 conda-forge
xdrfile 1.1.4 h7f98852_1 conda-forge
xtb 6.4.1 hf06ca72_2 conda-forge
xtb-python 20.2 py38h497a2fe_4 conda-forge
xz 5.2.5 h516909a_1 conda-forge
yaml 0.2.5 h7f98852_2 conda-forge
zipp 3.8.0 pyhd8ed1ab_0 conda-forge
zlib 1.2.11 h166bdaf_1014 conda-forge
Thanks, for the info. You are mixing the psi4 and conda-forge channel, which is currently not completely supported, if I understood Lori @loriab correctly. The psi4 channel is currently based on the Anaconda main channel (aka defaults), which is not necessarily compatible with conda-forge. I think one of the issues stems from having libgomp and intel-openmp in the same environment.
Good catch! This is what I get for wanting to use all the world's codes...
Indeed, it looks like using a minimalist conda env (which I thought I had done when testing this but clearly didn't) resolves the issue there. Thanks again for your help, @awvwgk! You're the best.
Note for the reader: This ended up being unrelated to the original issue raised.
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs.
![stale[bot] avatar](https://avatars.githubusercontent.com/in/1724?v=4 "Published by a pub.dev verified publisher"){kind=small}