-
Notifications
You must be signed in to change notification settings - Fork 10
/
utils.py
250 lines (209 loc) · 8.12 KB
/
utils.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
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
from datetime import datetime
import json
import logging
from pathlib import Path
import shutil
import sys
import tempfile
from xml.etree import ElementTree
import cophi
import flask
import numpy as np
import pandas as pd
from werkzeug.utils import secure_filename
from application import database
TEMPDIR = tempfile.gettempdir()
DATABASE = Path(TEMPDIR, "topicsexplorer.db")
LOGFILE = Path(TEMPDIR, "topicsexplorer.log")
DATA_EXPORT = Path(TEMPDIR, "topicsexplorer-data")
def init_app(name):
"""Initialize Flask application.
"""
logging.debug("Initializing Flask app...")
if getattr(sys, "frozen", False):
logging.debug("Application is frozen.")
root = Path(sys._MEIPASS)
else:
logging.debug("Application is not frozen.")
root = Path("application")
app = flask.Flask(name,
template_folder=str(Path(root, "templates")),
static_folder=str(Path(root, "static")))
return app
def init_logging(level):
"""Initialize logging.
"""
logging.basicConfig(level=level,
format="%(message)s",
filename=str(LOGFILE),
filemode="w")
# Disable logging for Flask and Werkzeug
# (this would be a lot of spam, even level INFO):
if level > logging.DEBUG:
logging.getLogger("flask").setLevel(logging.ERROR)
logging.getLogger("werkzeug").setLevel(logging.ERROR)
def init_db(app):
"""Initialize SQLite database.
"""
logging.debug("Initializing database...")
db = database.get_db()
if getattr(sys, "frozen", False):
root = Path(sys._MEIPASS)
else:
root = Path(".")
with app.open_resource(str(Path(root, "schema.sql"))) as schemafile:
schema = schemafile.read().decode("utf-8")
db.executescript(schema)
db.commit()
database.close_db()
def format_logging(message):
"""Format log messages.
"""
if "n_documents" in message:
n = message.split("n_documents: ")[1]
return "Number of documents: {}".format(n)
elif "vocab_size" in message:
n = message.split("vocab_size: ")[1]
return "Number of types: {}".format(n)
elif "n_words" in message:
n = message.split("n_words: ")[1]
return "Number of tokens: {}".format(n)
elif "n_topics" in message:
n = message.split("n_topics: ")[1]
return "Number of topics: {}".format(n)
elif "n_iter" in message:
return "Initializing topic model..."
elif "log likelihood" in message:
iteration, _ = message.split("> log likelihood: ")
return "Iteration {}".format(iteration[1:])
else:
return message
def load_textfile(textfile):
"""Load text file, return title and content.
"""
filename = Path(secure_filename(textfile.filename))
title = filename.stem
suffix = filename.suffix
if suffix in {".txt", ".xml", ".html"}:
content = textfile.read().decode("utf-8")
if suffix in {".xml", ".html"}:
content = remove_markup(content)
return title, content
# If suffix not allowed, ignore file:
else:
return None, None
def remove_markup(text):
"""Parse XML and drop tags.
"""
logging.info("Removing markup...")
tree = ElementTree.fromstring(text)
plaintext = ElementTree.tostring(tree,
encoding="utf8",
method="text")
return plaintext.decode("utf-8")
def get_documents(textfiles):
"""Get Document objects.
"""
logging.info("Processing documents...")
for textfile in textfiles:
title, content = textfile
yield cophi.model.Document(content, title)
def get_stopwords(data, corpus):
"""Get stopwords from file or corpus.
"""
logging.info("Fetching stopwords...")
if "stopwords" in data:
_, stopwords = load_textfile(data["stopwords"])
stopwords = cophi.model.Document(stopwords).tokens
else:
stopwords = corpus.mfw(data["mfw"])
return stopwords
def get_data(corpus, topics, iterations, stopwords, mfw):
"""Get data from HTML forms.
"""
logging.info("Processing user data...")
data = {"corpus": flask.request.files.getlist("corpus"),
"topics": int(flask.request.form["topics"]),
"iterations": int(flask.request.form["iterations"])}
if flask.request.files.get("stopwords", None):
data["stopwords"] = flask.request.files["stopwords"]
else:
data["mfw"] = int(flask.request.form["mfw"])
return data
def get_topics(model, vocabulary, maximum=100):
"""Get topics from topic model.
"""
logging.info("Fetching topics from topic model...")
for distribution in model.topic_word_:
words = list(np.array(vocabulary)[np.argsort(distribution)][:-maximum-1:-1])
yield "{}, ...".format(", ".join(words[:3])), words
def get_document_topic(model, titles, descriptors):
"""Get document-topic distribution from topic model.
"""
logging.info("Fetching document-topic distributions from topic model...")
document_topic = pd.DataFrame(model.doc_topic_)
document_topic.index = titles
document_topic.columns = descriptors
return document_topic
def get_cosine(matrix, descriptors):
"""Calculate cosine similarity between columns.
"""
logging.info("Calculcating cosine similarity...")
d = matrix.T @ matrix
norm = (matrix * matrix).sum(0, keepdims=True) ** .5
similarities = d / norm / norm.T
return pd.DataFrame(similarities, index=descriptors, columns=descriptors)
def scale(vector, minimum=50, maximum=100):
"""Min-max scaler for a vector.
"""
logging.debug("Scaling data from {} to {}...".format(minimum, maximum))
return np.interp(vector, (vector.min(), vector.max()), (minimum, maximum))
def export_data():
"""Export model output to ZIP archive.
"""
logging.info("Creating data archive...")
if DATA_EXPORT.exists():
unlink_content(DATA_EXPORT)
else:
DATA_EXPORT.mkdir()
model, stopwords = database.select("data_export")
document_topic, topics, document_similarities, topic_similarities = model
logging.info("Preparing document-topic distributions...")
document_topic = pd.read_json(document_topic, orient="index")
document_topic.columns = [col.replace(",", "").replace(" ...", "") for col in document_topic.columns]
logging.info("Preparing topics...")
topics = pd.read_json(topics, orient="index")
topics.index = ["Topic {}".format(n) for n in range(topics.shape[0])]
topics.columns = ["Word {}".format(n) for n in range(topics.shape[1])]
logging.info("Preparing topic similarity matrix...")
topic_similarities = pd.read_json(topic_similarities)
topic_similarities.columns = [col.replace(",", "").replace(" ...", "") for col in topic_similarities.columns]
topic_similarities.index = [ix.replace(",", "").replace(" ...", "") for ix in topic_similarities.index]
logging.info("Preparing document similarity matrix...")
document_similarities = pd.read_json(document_similarities)
data_export = {"document-topic-distribution": document_topic,
"topics": topics,
"topic-similarities": topic_similarities,
"document-similarities": document_similarities,
"stopwords": json.loads(stopwords)}
for name, data in data_export.items():
if name in {"stopwords"}:
with Path(DATA_EXPORT, "{}.txt".format(name)).open("w", encoding="utf-8") as file:
for word in data:
file.write("{}\n".format(word))
else:
path = Path(DATA_EXPORT, "{}.csv".format(name))
data.to_csv(path, sep=";", encoding="utf-8")
shutil.make_archive(DATA_EXPORT, "zip", DATA_EXPORT)
def unlink_content(directory, pattern="*"):
"""Deletes the content of a directory.
"""
logging.info("Cleaning up in data directory...")
for p in directory.rglob(pattern):
if p.is_file():
p.unlink()
def series2array(s):
"""Convert pandas Series to a 2-D array.
"""
for i, v in zip(s.index, s):
yield [i, v]