Open gelzinyte opened 3 years ago
hm... the XYZ formatting is in a bit of a flux right now. But some form of complex data type will be there (check out the #extxyz channel on our slack), so maybe we need to teach gap_fit to just treat that as a string and not worry about it.
how does gap_fit fail? there are already random string items in the xyz files that I use and they are fine. maybe it's the quoted quotes?
Not sure if it's just the quoted quotes (maybe non-escaped {
?), but it looks like the parsing of the info line gets messed up
$ head train.xyz
9
Lattice="50.0 0.0 0.0 0.0 50.0 0.0 0.0 0.0 50.0" Properties=species:S:1:pos:R:3:dft_forces:R:3 normal_mode_temperature=T dft_energy=-4212.509602197413 compound=ethanol n_atoms=9 config_type=ethanol_mol hash=50b506dbeef00fee5a23aceeb7ac0eb9 username=eg475 filename=train.xyz hash_structure=0154021c79357894b4136fd608fefad4 volume=124999.99999999991 mol_or_rad=mol dataset_type=train_compounds_train_configs normal_mode_energy=0.23389851991840674 formula=C2H6O elements="_JSON {\"6\": 2, \"8\": 1, \"1\": 6}" pbc="F F F"
C -0.93401342 0.07042787 0.05111524 0.35621935 -0.69215077 -0.84583725
C 0.53526377 -0.43666892 -0.11341624 -1.69537623 1.01342896 1.25964530
O 1.47668645 0.40662443 0.51842773 0.09388185 -0.42907879 0.50304526
$ bash sub.sh
Wed 14 Apr 17:27:59 BST 2021
SYSTEM ABORT: Traceback (most recent call last)
File "/opt/womble/QUIP/git_repo/src/libAtoms/xyz.c", line 765 kind IO
Missing value for parameter "6 pbc"
yes, it looks like the fortran parser doesn’t cope with the escaped quotes.
-- Gábor
Gábor Csányi Professor of Molecular Modelling Engineering Laboratory, University of Cambridge Pembroke College Cambridge
Pembroke College supports CARA. A Lifeline to Academics at Risk. http://www.cara.ngo/
On 14 Apr 2021, at 17:51, gelzinyte @.***> wrote:
Not sure if it's just the quoted quotes (maybe non-escaped {?), but it looks like the parsing of the info line gets messed up
$ head train.xyz 9 Lattice="50.0 0.0 0.0 0.0 50.0 0.0 0.0 0.0 50.0" Properties=species:S:1:pos:R:3:dft_forces:R:3 normal_mode_temperature=T dft_energy=-4212.509602197413 compound=ethanol n_atoms=9 config_type=ethanol_mol hash=50b506dbeef00fee5a23aceeb7ac0eb9 username=eg475 filename=train.xyz hash_structure=0154021c79357894b4136fd608fefad4 volume=124999.99999999991 mol_or_rad=mol dataset_type=train_compounds_train_configs normal_mode_energy=0.23389851991840674 formula=C2H6O elements="_JSON {\"6\": 2, \"8\": 1, \"1\": 6}" pbc="F F F" C -0.93401342 0.07042787 0.05111524 0.35621935 -0.69215077 -0.84583725 C 0.53526377 -0.43666892 -0.11341624 -1.69537623 1.01342896 1.25964530 O 1.47668645 0.40662443 0.51842773 0.09388185 -0.42907879 0.50304526
$ bash sub.sh Wed 14 Apr 17:27:59 BST 2021 SYSTEM ABORT: Traceback (most recent call last) File "/opt/womble/QUIP/git_repo/src/libAtoms/xyz.c", line 765 kind IO Missing value for parameter "6 pbc"
— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.
xyz frames from
abcd download
contain anelements
entry (e.g.elements="_JSON {\"6\": 2, \"8\": 1, \"1\": 6}"
) which messes with reading structures ingap_fit
. The simplest solution might be to just not writeelements
when abcd-downloading? Especially because there's also aformula
entry (formula=C2H6O
).