Initial commit of the classifuing code

classify4me_prototypes.py

+#! /usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+Includes function that reads json files and classifies the stuctures 
+included on the basis of encyclopedia and the calculated in the preprocessor 
+space_group and normalized_wyckoff, and returns a dictionary with key - json
+ file name and value - classification_name (with labels in the same format as 
+in the read_prototypes function).
+"""
+
+from __future__ import absolute_import
+
+__author__ = "Daria M. Tomecka"
+__copyright__ = "Copyright 2017, The NOMAD Project"
+__maintainer__ = "Daria M. Tomecka"
+__email__ = "tomeckadm@gmail.com;"
+__date__ = "18/05/17"
+
+import soap
+import sys
+import ase.io
+import json
+import numpy as np
+import pickle
+import h5py
+import time
+import datetime
+import os
+import logging
+
+from nomad_sim.wrappers import get_json_list
+from nomad_sim.wrappers import plot, logger
+from nomad_sim.utils_crystals import get_spacegroup
+from nomad_sim.utils_crystals import create_supercell
+from nomad_sim.wrappers import get_json_list
+from nomad_sim.nomad_structures import NOMADStructure
+from soap.tools import AseConfig
+
+
+#prototypes_file = os.path.normpath("/nomad-lab-base/analysis-tools/structural-similarity/python-modules/nomad_sim/structure_types.py")
+#Classification of json calculartion using encyclopedia preprocessor
+
+import sys
+#print(sys.path)
+sys.path.append('/home/beaker/py3k/lib/python3.5/')
+
+import Nomad
+import logging
+from Nomad.Preprocessing.System import structure, preprocessormaterial3d
+from Nomad import ProcessingEngine
+
+#LOGGER = logging.getLogger(__name__)
+#logging.basicConfig(level=logging.ERROR)
+
+json_list = beaker.json_list
+
+
+def classify_by_norm_wyckoff(json_list):
+    protos = {}
+    for json_file_name in json_list:
+        try:  
+            #try:
+            #    simulation_cell = json_file_name.get('lattice_vectors')
+            #    labels = json_file_name.get('atom_labels')
+            #    atom_pos = json_file_name.get('atom_positions')
+            #    if not simulation_cell or not labels or not atom_pos:
+            #        raise Exception("Parsed invalid geometry for space group %d, %s" % (spaceGroupNr, protoDict))
+
+                proc = ProcessingEngine(mode="debug")
+                results = proc.process(json_file_name)
+
+                preprocessor = list(results.values())
+                if len(preprocessor) == 1:
+                    preprocessor = preprocessor[0]
+                    spg_nr = preprocessor["space_group"]
+                    norm_wyckoff = preprocessor["normalized_wyckoff"]
+                    proto2 = preprocessor["structure_type"]
+                    protoDict = structure.get_structure_type(spg_nr, norm_wyckoff)
+                    #proto = '%d-%s' % (spg_nr, proto2)
+                    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' % (spg_nr, protoDict.get("Prototype","-"))
+                    protos[json_file_name] = proto
+                else:
+                    logging.error("failed for %r",json_file_name)
+        except:
+            logging.exception("failure while computing for %r",json_file_name)
+            
+           
+    return protos
+
+#classify_by_norm_wyckoff(json_list)
+