small pymatgen StructureMatcher test

pymatgen_matcher_test.py

from os.path import isfile
from pymatgen import Structure
from pymatgen.analysis.structure_matcher import StructureMatcher, ElementComparator
from pymatgen.io.cifio import CifWriter, CifParser

from mpworks.snl_utils.snl_mongo import SNLMongoAdapter
sma = SNLMongoAdapter.auto_load()

def get_structure(gid):
    ciffile = 'snlgroup_id_%d.cif' % gid
    if not isfile(ciffile):
        s = Structure.from_dict(sma.snlgroups.find_one({"snlgroup_id": gid})['canonical_snl'])
        w = CifWriter(s)
        w.write_file(ciffile)
    else:
        s = CifParser(ciffile).get_structures()[0]
    return s

# two sets of snlgroup_id pairs: a couple of canonical structures from the
# A. ~405 that didn't match a long time ago but now should match
# B. ~3500 that seemed like they matched a few months ago, but we realize now it was due to a bug
# duplicate check needs to return True for A. and False for B.!
setA = [
    [124070, 124069], [88026, 88025], [33468, 33466],
    [150493, 150494], [31564, 31563], [87801, 87800]
] 
setB = [
    [5335, 5336], [123806, 123805], [90125, 90124],
    [18884, 18883], [129197, 129196], [164662, 164639],
]

sm = StructureMatcher(
    ltol=0.2, stol=0.3, angle_tol=5, primitive_cell=True, scale=True,
    attempt_supercell=False, comparator=ElementComparator()
)

for list_of_gid_pairs in [setA, setB]:
    print list_of_gid_pairs
    for gid_pair in list_of_gid_pairs:
        s1, s2 = [get_structure(gid) for gid in gid_pair]
        print gid_pair, sm.fit(s1, s2)

pymatgen_matcher_unit_tests_cifs.tar.gz