-
Notifications
You must be signed in to change notification settings - Fork 0
/
prepare_llda_from_noaa.py
217 lines (169 loc) · 7.75 KB
/
prepare_llda_from_noaa.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
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
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
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
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
204
205
206
207
208
209
210
211
212
213
214
215
216
"""
prepare llda data from noaa data
"""
import os
import json
import sys
import re
import argparse
import codecs
import csv
from collections import namedtuple
#from myStemmer import pstem as stem
from myUtility.misc import Stopword_Handler
from stemming.porter2 import stem
Sinlge_LLDA_Document = namedtuple('Sinlge_LLDA_Document', ['did', 'labels','text'])
def get_data_from_cell(row,field_name):
cell_data = row[field_name.upper()]
cell_data = re.sub("\n"," ",cell_data)
cell_data = re.sub("\""," ",cell_data)
return cell_data
def process_event_type_string(event_type_string):
#origin = event_type_string
event_type_string = re.sub("/ ","/",event_type_string.lower())
event_type_string = re.sub(", ?","/",event_type_string)
event_type_string = re.sub("tstm","thunderstorm",event_type_string)
event_types = event_type_string.split("/")
for i in range(len(event_types)):
event_types[i] = re.sub("\s","_",event_types[i])
return event_types
def read_noaa_data(noaa_dir,use_disaster_cate,document_type):
if document_type == 0:
return read_data_use_event(noaa_dir,use_disaster_cate)
elif document_type == 1:
return read_data_use_episode(noaa_dir,use_disaster_cate)
else:
raise NotImplementedError("The document type %d is not implemented" %(document_type) )
def read_data_use_event(noaa_dir,use_disaster_cate):
# event_data = []
# episode_data = []
noaa_data = []
episode_ids = {}
for year_csv_file in os.walk(noaa_dir).next()[2]:
year_csv_file = os.path.join(noaa_dir,year_csv_file)
print year_csv_file
with open(year_csv_file,"rb") as f:
reader = csv.DictReader(f)
for row in reader:
episode_id = get_data_from_cell(row,"episode_id")
event_id = get_data_from_cell(row,"event_id")
event_type_string = get_data_from_cell(row,"event_type")
event_types = process_event_type_string(event_type_string)
episode_narrative = get_data_from_cell(row,"episode_narrative")
event_narrative = get_data_from_cell(row,"event_narrative")
if episode_id not in episode_ids:
if len(episode_narrative) != 0:
episode_ids[episode_id] = 0
did = episode_id
labels = " ".join(event_types+[episode_id])
if use_disaster_cate:
labels += " %s" %("disaster")
text = episode_narrative
noaa_data.append( Sinlge_LLDA_Document._make([did,labels,text]) )
# episode_data.append( Sinlge_LLDA_Document._make(did,labels,text) )
if len(event_narrative)==0:
continue
did = "%s_%s" %(episode_id,event_id)
labels = " ".join(event_types+[episode_id])
if use_disaster_cate:
labels += " %s" %("disaster")
text = event_narrative
noaa_data.append( Sinlge_LLDA_Document._make([did,labels,text]) )
# event_data.append( Sinlge_LLDA_Document._make(did,labels,text) )
# print "There are %d events and %d episodes" %(len(event_data),len(episode_data))
print "There are %d documents" %(len(noaa_data))
return noaa_data
def read_data_use_episode(noaa_dir,use_disaster_cate):
episode_data = {}
all_lables = set()
for year_csv_file in os.walk(noaa_dir).next()[2]:
year_csv_file = os.path.join(noaa_dir,year_csv_file)
print year_csv_file
with open(year_csv_file,"rb") as f:
reader = csv.DictReader(f)
for row in reader:
episode_id = get_data_from_cell(row,"episode_id")
event_id = get_data_from_cell(row,"event_id")
event_type_string = get_data_from_cell(row,"event_type")
event_types = process_event_type_string(event_type_string)
for e_type in event_types:
all_lables.add(e_type)
episode_narrative = get_data_from_cell(row,"episode_narrative")
event_narrative = get_data_from_cell(row,"event_narrative")
if len(episode_narrative) == 0 and len(event_narrative)==0:
continue
if not episode_id.isdigit():
continue
if episode_id not in episode_data:
episode_data[episode_id]= {
"labels":event_types,
"episode":episode_narrative,
"event":event_narrative
}
else:
if len(episode_narrative) != 0 and len(episode_data[episode_id]["episode"]) == 0:
episode_data[episode_id]["episode"] = episode_narrative
if len(event_narrative) != 0:
episode_data[episode_id]["event"] += " %s" %(event_narrative)
for e_type in event_types:
if e_type not in episode_data[episode_id]["labels"]:
episode_data[episode_id]["labels"].append(e_type)
noaa_data = []
for episode_id in episode_data:
did = episode_id
labels = " ".join(episode_data[episode_id]["labels"])
if use_disaster_cate:
labels += " %s" %("disaster")
text = "%s %s" %(episode_data[episode_id]["episode"],
episode_data[episode_id]["event"])
noaa_data.append( Sinlge_LLDA_Document._make([did,labels,text]) )
# event_data.append( Sinlge_LLDA_Document._make(did,labels,text) )
# print "There are %d events and %d episodes" %(len(event_data),len(episode_data))
print "There are %d documents" %(len(noaa_data))
print all_lables
return noaa_data
def write_llda_input(noaa_data, dest_file,process,stopword_handler):
with open(dest_file,'wb') as f:
spamwriter = csv.writer(f)
for single_document in noaa_data:
if process:
document_text = stopword_handler.remove_stopwords(single_document.text.lower())
new_text = ""
for term in re.findall("\w+",document_text):
new_text += " %s" %(stem(term))
spamwriter.writerow(
[
single_document.did,
single_document.labels,
new_text
]
)
else:
spamwriter.writerow(
[
single_document.did,
single_document.labels,
single_document.text
]
)
def main():
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument("noaa_dir")
parser.add_argument("dest_file")
parser.add_argument("--use_disaster_cate","-ud",action='store_true')
parser.add_argument("--process","-p",action='store_true')
parser.add_argument("--document_type","-dt",type=int, choices=range(2),default=0,
help="""
Chose the type of document that will be used:
0:event
1:episode
""")
args=parser.parse_args()
if args.process:
stopword_handler = Stopword_Handler()
else:
stopword_handler = None
noaa_data = read_noaa_data(args.noaa_dir,args.use_disaster_cate,args.document_type)
write_llda_input(noaa_data, args.dest_file,args.process,stopword_handler)
if __name__=="__main__":
main()