classify4me_prototypes.py 8.89 KB
Newer Older
1
2
3
#! /usr/bin/env python3
# -*- coding: utf-8 -*-
"""
4
5
6
Reads calculation data and classifies their structures by prototypes 
on the basis of the space_group and normalized_wyckoff, and adds labels to the calculations
 - prototype_label (with labels in the same format as 
7
8
9
10
11
12
13
14
15
in the read_prototypes function).
"""

from __future__ import absolute_import

__author__ = "Daria M. Tomecka and Fawzi Mohamed"
__copyright__ = "Copyright 2017, The NOMAD Project"
__maintainer__ = "Daria M. Tomecka"
__email__ = "tomeckadm@gmail.com;"
16
__date__ = "12/07/17"
17
18
19

import sys
import ase.io
20
from ase.data import chemical_symbols
21
22
23
24
25
26
import json
import numpy as np
import time
import datetime
import os, os.path
import logging
27
28
import functools
import fractions
29
30

import setup_paths
31
32

### new updated location for the prototypes
33
34
35
36
37
38
39
40
41
42
43
44
45
from nomadcore.structure_types import \
    structure_types_by_spacegroup as str_types_by_spg
from nomadcore.parse_streamed_dicts import ParseStreamedDicts
from nomadcore.local_meta_info import loadJsonFile, InfoKindEl
from nomadcore.parser_backend import JsonParseEventsWriterBackend


#prototypes_file = os.path.normpath("/nomad-lab-base/analysis-tools/structural-similarity/python-modules/nomad_sim/structure_types.py")

import sys
#print(sys.path)
sys.path.append('/home/beaker/py3k/lib/python3.5/')
import spglib
46

47
48
49
50
51
52
53
import logging

#LOGGER = logging.getLogger(__name__)
#logging.basicConfig(level=logging.ERROR)


atomSpecies = None
54

55
cell = None
56
57
58
59
60
61
62
63
64
65

def get_normalized_wyckoff(atomic_number, wyckoff):
    """Returns a normalized Wyckoff sequence for the given atomic numbers and
    wyckoff symbols.
    """
    # print("at", atomic_number, wyckoff)
    atomCount = {}
    for nr in atomic_number:
        atomCount[nr] = atomCount.get(nr, 0) + 1
    wycDict = {}
66
    #logging.error("atomic_number: %s, wyckoff: %s", atomic_number, wyckoff) 
67
68
    for i, wk in enumerate(wyckoff):
        oldVal = wycDict.get(wk, {})
69
        #print("i:",i, "wyckoff", wyckoff, "wk", wk)
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
        nr = atomic_number[i]
        oldVal[nr] = oldVal.get(nr, 0) + 1
        wycDict[wk] = oldVal
    sortedWyc = list(wycDict.keys())
    sortedWyc.sort()

    def cmpp(a, b):
        return ((a < b) - (a > b))

    def compareAtNr(at1, at2):
        """Returns a normalized Wyckoff sequence for the given atomic numbers and
        wyckoff symbols.
        """
        # print("at", atomic_number, wyckoff)
        atomCount = {}
        for nr in atomic_number:
            atomCount[nr] = atomCount.get(nr, 0) + 1
        wycDict = {}
        for i, wk in enumerate(wyckoff):
            oldVal = wycDict.get(wk, {})
            nr = atomic_number[i]
            oldVal[nr] = oldVal.get(nr, 0) + 1
            wycDict[wk] = oldVal
        sortedWyc = list(wycDict.keys())
        sortedWyc.sort()

    def cmpp(a, b):
        return ((a < b) - (a > b))

    def compareAtNr(at1, at2):
        c = cmpp(atomCount[at1], atomCount[at2])
        if (c != 0):
            return c
        for wk in sortedWyc:
            p = wycDict[wk]
            c = cmpp(p.get(at1, 0), p.get(at2, 0))
            if c != 0:
                return c
        return 0
109

110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
    sortedAt = list(atomCount.keys())
    sortedAt.sort(key=functools.cmp_to_key(compareAtNr))
    standardAtomNames = {}
    for i, at in enumerate(sortedAt):
        standardAtomNames[at] = ("X_%d" % i)
    standardWyc = {}
    for wk, ats in wycDict.items():
        stdAts = {}
        for at, count in ats.items():
            stdAts[standardAtomNames[at]] = count
        standardWyc[wk] = stdAts
    if standardWyc:
        counts = [c for x in standardWyc.values() for c in x.values()]
        # logging.error("counts: %s", counts)
        gcd = counts[0]
        for c in counts[1:]:
            gcd = fractions.gcd(gcd,c)
        if gcd != 1:
            for wk,d in standardWyc.items():
                for at,c in d.items():
                    d[at] = c // gcd
    return standardWyc



def get_structure_type(space_group, norm_wyckoff):
    """Returns the information on the prototype.
    """
    structure_type_info = {}

    for type_description in str_types_by_spg.get((space_group), []):
141
        
142
        current_norm_wyckoffs = type_description.get("normalized_wyckoff_spg")
143
144
145
146
147
148
149
150
151
152
153
154
        if current_norm_wyckoffs and current_norm_wyckoffs == norm_wyckoff:
            structure_type_info = type_description
            break
    if structure_type_info:
        return structure_type_info
    else:
        return None



def _structure_type_info(self):
    """Known structure types"""
155
    return get_structure_type(
156
157
158
        self.space_group,
        self.normalized_wyckoff)

159

160
def toAtomNr(string):
161
    "returns the atom number of the given symbol"
162
    baseStr = string[:3].title()
163
164
165
166
167
168
169
170
171
    if baseStr.startswith("Uu") and baseStr in chemical_symbols[1:]:
        return chemical_symbols.index(baseStr)
    if baseStr[:2] in chemical_symbols[1:]:
        return chemical_symbols.index(baseStr[:2])
    elif baseStr[:1] in chemical_symbols[1:]:
        return chemical_symbols.index(baseStr[:1])
    else:
        return 0

172
def dictToNarray(dictValue):
173
    """function that gets the dictionary with flat data ans shape and give the array"""
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
    v=dictValue['flatData']
    return np.reshape(np.asarray(v), dictValue['shape'])

def protoNormalizeWycoff(protoDict):
    """recalculates the normalized wyckoff values for the given prototype dictionary"""
    cell = np.asarray(protoDict['lattice_vectors'])
    atomSpecies = [toAtomNr(at) for at in protoDict['atom_labels']]
    atomPos = np.asarray(protoDict['atom_positions'])
    symm = systemToSpg(cell, atomSpecies, atomPos)
    wyckoffs = symm.get("wyckoffs")
    norm_wyckoff = get_normalized_wyckoff(atomSpecies,wyckoffs) 
    return norm_wyckoff

def updatePrototypesWyckoff(protos):
    for sp, pts in protos.items():
        for protoDict in pts:
            try:
                wy = protoNormalizeWycoff(protoDict)
                protoDict['normalized_wyckoff_spg'] = wy
            except:
                logging.exception("Failed to compute normalized wyckoffs for %s", protoDict)

def systemToSpg(cell, atomSpecies, atomPos):
    """uses spg to calculate the symmetry of the given system"""
    acell = cell*1.0e10
    cellInv = np.linalg.inv(cell)
    symm = spglib.get_symmetry_dataset((acell, np.dot(atomPos,cellInv), atomSpecies),
                                                 0.002, -1) # use m instead of Angstrom?

    return symm
204
205
206

def classify_by_norm_wyckoff(sectionSystem):
    try:
207
        
208
209
210
        ###
        #atomic_number = atom_species
        #as in the normalized version
211
212
213
214
215
216
217
218
219
220
221
222
223
        cell = None
        conf = sectionSystem
        lab = conf.get("atom_labels", None)
        ##periodicDirs = conf.get("configuration_periodic_dimensions", periodicDirs)
        atomSpecies = [toAtomNr(l) for l in lab[0]['flatData']]
        #print (atomSpecies)
        newCell = conf.get("simulation_cell")
        if newCell:
            cell = dictToNarray(newCell)
        symm = None
        #print("***full:",cell)
        #acell = cell.reshape(3,3)
        atomPos = dictToNarray(conf.get("atom_positions")[0])
224

225
226
227
        symm = systemToSpg(cell, atomSpecies, atomPos)
        wyckoffs = symm.get("wyckoffs")
        spg_nr = symm.get("number")
228
229
        
        ### adds recalculated wyckoff positions
230
        updatePrototypesWyckoff(str_types_by_spg)
231
232
233

        ###
        norm_wyckoff = get_normalized_wyckoff(atomSpecies,wyckoffs) 
234
        protoDict = get_structure_type(spg_nr, norm_wyckoff)
235
236
237
238
239
240
241
242
243
244
245

        if protoDict is None:
            proto = "%d-_" % spg_nr
        else:
            #if protoDict.get("Notes","") not in ["", "_", "-", "–"]:
            #    proto = '%d-%s' % (spg_nr, protoDict)
            #else:
            #proto = '%d-%s' % (spg_nr, protoDict)
            proto = '%d-%s-%s' % (spg_nr, protoDict.get("Prototype","-"),protoDict.get("Pearsons Symbol","-"))
        return proto
    except:
246
247
        #logging.exception("failure while computing for %r",json_file_name)
        logging.exception("failure while computing for that example")
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271

    return None

#classify_by_norm_wyckoff(json_list)
            

def main():
    metapath = '../../../../nomad-meta-info/meta_info/nomad_meta_info/' +\
          'common.nomadmetainfo.json'
    metaInfoPath = os.path.normpath(
    os.path.join(os.path.dirname(os.path.abspath(__file__)), metapath))
  
    metaInfoEnv, warns = loadJsonFile(filePath=metaInfoPath,
                                      dependencyLoader=None,
                                      extraArgsHandling=InfoKindEl.ADD_EXTRA_ARGS,
                                      uri=None)
    backend = JsonParseEventsWriterBackend(metaInfoEnv)
    dictReader = ParseStreamedDicts(sys.stdin)
    while True:
        sectSys = dictReader.readNextDict()
        if sectSys is None:
            break
        label = classify_by_norm_wyckoff(sectSys)
        if label:
272
            backend.openContext(sectSys['uri'])
273
            backend.addValue("prototype_label", label)
274
            backend.closeContext(sectSys['uri'])
275
276
277
278

if __name__ == '__main__':
    main()