nielstron / quantulum3 / 946

pending completion

Build # 946

Build Type

cron

travis-ci-com

Committed by

nielstron

Commit Message

Merge branch 'dev'

Run Details

467 of 467 new or added lines in 14 files covered. (100.0%)

1812 of 1847 relevant lines covered (98.11%)

4.89 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

97.06

/quantulum3/scripts/train.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-
""" Train the disambiguator """

import argparse
import json
import logging
import time

from quantulum3.classifier import train_classifier

_LOGGER = logging.getLogger(__name__)


def main(args=None):
    logging.basicConfig(level=logging.INFO)
    parser = argparse.ArgumentParser(
        "train",
        description="Train unit disambiguator based on data in quantulum "
        "project folder",
    )
    parser.add_argument(
        "--store",
        "-s",
        help="store resulting classifier as a joblib file",
        action="store_true",
    )
    parser.add_argument(
        "--lang",
        "-l",
        help="language in which to train the classifier, default 'en_US'",
        type=str,
        default="en_US",
    )
    parser.add_argument(
        "--data",
        "-d",
        help=(
            "path to one or more json training files. If none are given, the default training "
            "data in the quantulum3 project folder is used. Multiple files can be given by "
            "specifying this argument multiple times (e.g. -d train1.json -d train2.json). "
            "See the project README for more information."
        ),
        type=str,
        action="append",
    )
    parser.add_argument(
        "--output",
        "-o",
        help=(
            "path to a folder where the resulting classifier is stored. If "
            "not given, the default location in the quantulum3 project folder "
            "is used"
        ),
        type=str,
        default=None,
    )

    args = parser.parse_args(args)

    if args.data is not None:
        training_set = []
        args.lang = None
        for file in args.data:
            with open(file) as f:
                training_set += json.load(f)
    else:
        training_set = None

    if args.output is not None:
        # override this option if an output file is given, feels intuitive to do
        args.store = True

    _LOGGER.info(
        "Start training for language {}, {}storing the classifier".format(
            args.lang, "" if args.store else "not "
        )
    )
    start = time.process_time()
    try:
        train_classifier(
            store=args.store,
            lang=args.lang,
            training_set=training_set,
            output_path=args.output,
        )
    except ImportError:
        _LOGGER.error(
            "Could not train the classifier. Make sure you have the "
            f"required dependencies installed. "
            "These can be installed in pip using the command "
            "'pip install quantulum3[classifier]'"
        )
    else:
        end = time.process_time()
        _LOGGER.info("Done in {} s".format(end - start))


if __name__ == "__main__":
    main()

1	#!/usr/bin/env python
2	# -- coding: utf-8 --
3	""" Train the disambiguator """	5✔
4
5	import argparse	5✔
6	import json	5✔
7	import logging	5✔
8	import time	5✔
9
10	from quantulum3.classifier import train_classifier	5✔
11
12	_LOGGER = logging.getLogger(__name__)	5✔
13
14
15	def main(args=None):	5✔
16	logging.basicConfig(level=logging.INFO)	5✔
17	parser = argparse.ArgumentParser(	5✔
18	"train",
19	description="Train unit disambiguator based on data in quantulum "
20	"project folder",
21	)
22	parser.add_argument(	5✔
23	"--store",
24	"-s",
25	help="store resulting classifier as a joblib file",
26	action="store_true",
27	)
28	parser.add_argument(	5✔
29	"--lang",
30	"-l",
31	help="language in which to train the classifier, default 'en_US'",
32	type=str,
33	default="en_US",
34	)
35	parser.add_argument(	5✔
36	"--data",
37	"-d",
38	help=(
39	"path to one or more json training files. If none are given, the default training "
40	"data in the quantulum3 project folder is used. Multiple files can be given by "
41	"specifying this argument multiple times (e.g. -d train1.json -d train2.json). "
42	"See the project README for more information."
43	),
44	type=str,
45	action="append",
46	)
47	parser.add_argument(	5✔
48	"--output",
49	"-o",
50	help=(
51	"path to a folder where the resulting classifier is stored. If "
52	"not given, the default location in the quantulum3 project folder "
53	"is used"
54	),
55	type=str,
56	default=None,
57	)
58
59	args = parser.parse_args(args)	5✔
60
61	if args.data is not None:	5✔
62	training_set = []	5✔
63	args.lang = None	5✔
64	for file in args.data:	5✔
65	with open(file) as f:	5✔
66	training_set += json.load(f)	5✔
67	else:
68	training_set = None	5✔
69
70	if args.output is not None:	5✔
71	# override this option if an output file is given, feels intuitive to do
72	args.store = True	5✔
73
74	_LOGGER.info(	5✔
75	"Start training for language {}, {}storing the classifier".format(
76	args.lang, "" if args.store else "not "
77	)
78	)
79	start = time.process_time()	5✔
80	try:	5✔
81	train_classifier(	5✔
82	store=args.store,
83	lang=args.lang,
84	training_set=training_set,
85	output_path=args.output,
86	)
87	except ImportError:	5✔
88	_LOGGER.error(	5✔
89	"Could not train the classifier. Make sure you have the "
90	f"required dependencies installed. "
91	"These can be installed in pip using the command "
92	"'pip install quantulum3[classifier]'"
93	)
94	else:
95	end = time.process_time()	5✔
96	_LOGGER.info("Done in {} s".format(end - start))	5✔
97
98
99	if __name__ == "__main__":	5✔
100	main()	×

nielstron / quantulum3 / 946

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous