def from_config(cls, config, name, section_key="extractors"):
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
elif 'class' in section:
Class = yamlconf.import_module(section['class'])
return Class.from_config(config, name)
def from_config(cls, config, name, section_key="tokenizers"):
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
else:
Tokenizer = yamlconf.import_module(section['class'])
return Tokenizer.from_config(config, name, section_key)
def from_config(self, config, name, section_key="languages"):
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
elif 'class' in section:
Class = yamlconf.import_module(section['class'])
return Class.from_config(config, section_key)
def from_config(cls, config, name, section_key="extractors"):
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
elif 'class' in section:
Class = yamlconf.import_module(section['class'])
return Class.from_config(config, name)
def from_config(cls, config, name, section_key="score_caches"):
logger.info("Loading ScoreCache '{0}' from config.".format(name))
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
elif 'class' in section:
Class = yamlconf.import_module(section['class'])
return Class.from_config(config, name)
def from_config(cls, config, name, section_key='scorer_models'):
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
elif 'class' in section:
class_path = section['class']
Class = yamlconf.import_module(class_path)
assert cls != Class
return Class.from_config(config, name)
def from_config(cls, config, name, section_key="diff_engines"):
"""
Constructs a :class:`deltas.DiffEngine` from a configuration
doc.
"""
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
else:
Engine = yamlconf.import_module(section['class'])
return Engine.from_config(config, name, section_key=section_key)
def from_config(cls, config, name, section_key="diff_engines"):
"""
Constructs a :class:`deltas.DiffEngine` from a configuration
doc.
"""
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
else:
Engine = yamlconf.import_module(section['class'])
return Engine.from_config(config, name, section_key=section_key)
def from_config(cls, config, name, section_key='scorer_models'):
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
elif 'class' in section:
class_path = section['class']
Class = yamlconf.import_module(class_path)
assert cls != Class
return Class.from_config(config, name, section_key=section_key)
def main(argv=None):
args = docopt.docopt(__doc__, argv=argv)
logging.basicConfig(
level=logging.INFO if not args['--debug'] else logging.DEBUG,
format='%(asctime)s %(levelname)s:%(name)s -- %(message)s')
ScoringModel = yamlconf.import_module(args['<scoring-model>'])
features = yamlconf.import_module(args['<features>'])
version = args['--version']
estimator_params = {}
for parameter in args['--parameter']:
key, value = parameter.split("=", 1)
estimator_params[key] = json.loads(value)
labels, label_weights, population_rates = read_labels_and_population_rates(
args['--labels'], args['--label-weight'], args['--pop-rate'],
args['--labels-config'])
multilabel = False
if args['--multilabel']:
multilabel = True
model = ScoringModel(features,
version=version,
multilabel=multilabel,
labels=labels,
label_weights=label_weights,
population_rates=population_rates,
center=args['--center'],
scale=args['--scale'],
**estimator_params)
if args['--observations'] == "<stdin>":
observations = read_observations(sys.stdin)
else:
observations = read_observations(open(args['--observations']))
label_name = args['<label>']
value_labels = \
[(list(solve(features, cache=ob['cache'])), ob[label_name])
for ob in observations]
if args['--model-file'] == "<stdout>":
model_file = sys.stdout.buffer
else:
model_file = open(args['--model-file'], 'wb')
folds = int(args['--folds'])
workers = int(args['--workers']) if args['--workers'] is not None else None
run(value_labels, model_file, model, folds, workers)
def main(argv=None):
args = docopt.docopt(__doc__, argv=argv)
logging.basicConfig(
level=logging.INFO if not args['--debug'] else logging.DEBUG,
format='%(asctime)s %(levelname)s:%(name)s -- %(message)s'
)
ScoringModel = yamlconf.import_module(args['<scoring-model>'])
features = yamlconf.import_module(args['<features>'])
version = args['--version']
estimator_params = {}
for parameter in args['--parameter']:
key, value = parameter.split("=", 1)
estimator_params[key] = json.loads(value)
labels, label_weights, population_rates = read_labels_and_population_rates(
args['--labels'], args['--label-weight'], args['--pop-rate'],
args['--labels-config'])
multilabel = False
if args['--multilabel']:
multilabel = True
model = ScoringModel(
features, version=version, multilabel=multilabel,
labels=labels, label_weights=label_weights,
population_rates=population_rates,
center=args['--center'],
scale=args['--scale'],
**estimator_params)
if args['--observations'] == "<stdin>":
observations = read_observations(sys.stdin)
else:
observations = read_observations(open(args['--observations']))
label_name = args['<label>']
value_labels = \
[(list(solve(features, cache=ob['cache'])), ob[label_name])
for ob in observations]
if args['--model-file'] == "<stdout>":
model_file = sys.stdout.buffer
else:
model_file = open(args['--model-file'], 'wb')
folds = int(args['--folds'])
workers = int(args['--workers']) if args['--workers'] is not None else None
run(value_labels, model_file, model, folds, workers)
def from_config(cls, config, name, section_key="score_caches"):
try:
import yamlconf
except ImportError:
raise ImportError("Could not find yamlconf. This packages is " +
"required when using yaml config files.")
logger.info("Loading ScoreCache '{0}' from config.".format(name))
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
elif 'class' in section:
Class = yamlconf.import_module(section['class'])
return Class.from_config(config, name)
def from_config(cls, config, name, section_key="metrics_collectors"):
try:
import yamlconf
except ImportError:
raise ImportError("Could not find yamlconf. This packages is " +
"required when using yaml config files.")
logger.info("Loading MetricsCollector '{0}' from config.".format(name))
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
elif 'class' in section:
Class = yamlconf.import_module(section['class'])
return Class.from_config(config, name)
def main(argv=None):
args = docopt.docopt(__doc__, argv=argv)
logging.basicConfig(
level=logging.INFO if not args['--debug'] else logging.DEBUG,
format='%(asctime)s %(levelname)s:%(name)s -- %(message)s')
sys.path.insert(0, ".") # Search local directory first
ScorerModel = yamlconf.import_module(args['<scorer-model>'])
features = yamlconf.import_module(args['<features>'])
version = args['--version']
estimator_params = {}
for parameter in args['--parameter']:
key, value = parameter.split("=")
estimator_params[key] = json.loads(value)
test_statistics = []
for stat_str in args['--statistic']:
test_statistics.append(TestStatistic.from_stat_str(stat_str))
scorer_model = ScorerModel(
features,
version=version,
balanced_sample=args['--balance-sample'],
balanced_sample_weight=args['--balance-sample-weight'],
center=args['--center'],
scale=args['--scale'],
**estimator_params)
if args['--observations'] == "<stdin>":
observations = read_observations(sys.stdin)
else:
observations = read_observations(open(args['--observations']))
label_name = args['<label>']
value_labels = \
[(list(solve(features, cache=ob['cache'])), ob[label_name])
for ob in observations]
if args['--model-file'] == "<stdout>":
model_file = sys.stdout.buffer
else:
model_file = open(args['--model-file'], 'wb')
folds = int(args['--folds'])
workers = int(args['--workers']) if args['--workers'] is not None else None
run(value_labels, model_file, scorer_model, test_statistics, folds,
workers)
def from_config(cls, config, name, section_key='scorer_models'):
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
elif 'class' in section:
class_path = section['class']
Class = yamlconf.import_module(class_path)
if 'model_file' in section:
return Class.load(open(section['model_file'], 'rb'))
else:
return Class(
**{k: v
for k, v in section.items() if k != "class"})
def main(argv=None):
args = docopt.docopt(__doc__, argv=argv)
logging.basicConfig(
level=logging.INFO if not args['--debug'] else logging.DEBUG,
format='%(asctime)s %(levelname)s:%(name)s -- %(message)s'
)
ScorerModel = yamlconf.import_module(args['<scorer_model>'])
features = yamlconf.import_module(args['<features>'])
version = args['--version']
model_kwargs = {}
for parameter in args['--parameter']:
key, value = parameter.split("=")
model_kwargs[key] = json.loads(value)
test_statistics = []
for stat_str in args['--statistic']:
test_statistics.append(TestStatistic.from_stat_str(stat_str))
scorer_model = ScorerModel(
features, version=version,
balanced_sample=args['--balance-sample'],
balanced_sample_weight=args['--balance-sample-weight'],
center=args['--center'],
scale=args['--scale'],
**model_kwargs)
if args['--values-labels'] == "<stdin>":
observations_f = sys.stdin
else:
observations_f = open(args['--values-labels'], 'r')
if args['--model-file'] == "<stdout>":
model_file = sys.stdout.buffer
else:
model_file = open(args['--model-file'], 'wb')
decode_label = util.DECODERS[args['--label-type']]
observations = util.read_observations(observations_f,
scorer_model.features,
decode_label)
test_prop = float(args['--test-prop'])
run(observations, model_file, scorer_model, test_statistics, test_prop)
def from_config(cls, config, name, section_key='scorer_models'):
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
elif 'class' in section:
class_path = section['class']
Class = yamlconf.import_module(class_path)
if 'model_file' in section:
# TODO: Cache the model file for reuse across workers?
with open_file(section['model_file']) as stream:
return Class.load(stream)
else:
return Class(**{k: v for k, v in section.items()
if k != "class"})
def from_config(cls, config, name, section_key='scorer_models'):
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
elif 'class' in section:
class_path = section['class']
Class = yamlconf.import_module(class_path)
if 'model_file' in section:
# TODO: Cache the model file for reuse across workers?
with open_file(section['model_file']) as stream:
return Class.load(stream)
else:
return Class(**{k: v for k, v in section.items()
if k != "class"})
def from_config(cls, config, name, section_key="scoring_systems"):
try:
import yamlconf
except ImportError:
raise ImportError("Could not find yamlconf. This packages is " +
"required when using yaml config files.")
logger.info("Loading ScoreProcessor '{0}' from config.".format(name))
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
elif 'class' in section:
Class = yamlconf.import_module(section['class'])
return Class.from_config(config, name)
else:
raise RuntimeError("No module or class to load.")
def main(argv=None):
args = docopt.docopt(__doc__, argv=argv)
logging.basicConfig(
level=logging.INFO if not args['--debug'] else logging.DEBUG,
format='%(asctime)s %(levelname)s:%(name)s -- %(message)s')
ScorerModel = yamlconf.import_module(args['<scorer_model>'])
features = yamlconf.import_module(args['<features>'])
version = args['--version']
model_kwargs = {}
for parameter in args['--parameter']:
key, value = parameter.split("=")
model_kwargs[key] = json.loads(value)
test_statistics = []
for stat_str in args['--statistic']:
test_statistics.append(TestStatistic.from_stat_str(stat_str))
scorer_model = ScorerModel(
features,
version=version,
balanced_sample=args['--balance-sample'],
balanced_sample_weight=args['--balance-sample-weight'],
center=args['--center'],
scale=args['--scale'],
**model_kwargs)
if args['--values-labels'] == "<stdin>":
observations_f = sys.stdin
else:
observations_f = open(args['--values-labels'], 'r')
if args['--model-file'] == "<stdout>":
model_file = sys.stdout.buffer
else:
model_file = open(args['--model-file'], 'wb')
decode_label = util.DECODERS[args['--label-type']]
observations = util.read_observations(observations_f,
scorer_model.features, decode_label)
test_prop = float(args['--test-prop'])
run(observations, model_file, scorer_model, test_statistics, test_prop)
def main(argv=None):
args = docopt.docopt(__doc__, argv=argv)
logging.basicConfig(
level=logging.WARNING if not args['--debug'] else logging.DEBUG,
format='%(asctime)s %(levelname)s:%(name)s -- %(message)s')
features = yamlconf.import_module(args['<features>'])
session = mwapi.Session(args['--host'],
user_agent="Revscoring feature extractor utility")
extractor = api.Extractor(session)
if args['--rev-labels'] == "<stdin>":
rev_labels = read_rev_labels(sys.stdin)
else:
rev_labels = read_rev_labels(open(args['--rev-labels']))
if args['--value-labels'] == "<stdout>":
value_labels = sys.stdout
else:
value_labels = open(args['--value-labels'], 'w')
include_revid = bool(args['--include-revid'])
if args['--extractors'] == "<cpu count>":
extractors = cpu_count()
else:
extractors = int(args['--extractors'])
verbose = args['--verbose']
debug = args['--debug']
run(rev_labels, value_labels, features, extractor, include_revid,
extractors, verbose, debug)
def main(argv=None):
args = docopt.docopt(__doc__, argv=argv)
logging.basicConfig(
level=logging.INFO if not args['--debug'] else logging.DEBUG,
format='%(asctime)s %(levelname)s:%(name)s -- %(message)s')
sys.path.insert(0, ".") # Search local directory first
features = yamlconf.import_module(args['<features>'])
label_name = args['<label>']
if args['<model>'] is not None:
model = Model.load(open(args['<model>']))
else:
model = None
additional_fields = args['<additional-field>']
if args['--input'] == "<stdin>":
observations = read_observations(sys.stdin)
else:
observations = read_observations(open(args['--input']))
if args['--output'] == "<stdout>":
output = sys.stdout
else:
output = open(args['--output'], 'w')
verbose = args['--verbose']
run(observations, output, features, label_name, model, additional_fields,
verbose)
def from_config(cls, config, name, section_key="segmenters"):
"""
Constructs a segmenter from a configuration doc.
"""
section = config[section_key][name]
segmenter_class_path = section['class']
Segmenter = yamlconf.import_module(segmenter_class_path)
return Segmenter.from_config(config, name, section_key=section_key)
def main(argv=None):
args = docopt.docopt(__doc__, argv=argv)
logging.basicConfig(
level=logging.INFO if not args['--debug'] else logging.DEBUG,
format='%(asctime)s %(levelname)s:%(name)s -- %(message)s'
)
observations = read_observations(sys.stdin)
sys.path.insert(0, ".") # Search local directory first
features = yamlconf.import_module(args['<features>'])
label_name = args['<label>']
verbose = args['--verbose']
run(observations, features, label_name, verbose)
def main(argv=None):
args = docopt.docopt(__doc__, argv=argv)
logging.basicConfig(
level=logging.INFO if not args['--debug'] else logging.DEBUG,
format='%(asctime)s %(levelname)s:%(name)s -- %(message)s'
)
params_config = yamlconf.load(open(args['<params-config>']))
features_path = args['<features>']
features = yamlconf.import_module(features_path)
label_decoder = util.DECODERS[args['--label-type']]
if args['--observations'] == "<stdin>":
observations_f = sys.stdin
else:
observations_f = open(args['--observations'])
observations = util.read_observations(observations_f, features,
label_decoder)
# Get a sepecialized scorer if we have one
scoring = metrics.SCORERS.get(args['--scoring'], args['--scoring'])
folds = int(args['--folds'])
if args['--report'] == "<stdout>":
report = sys.stdout
else:
report = open(args['--report'], "w")
if args['--processes'] == "<cpu-count>":
processes = multiprocessing.cpu_count()
else:
processes = int(args['--processes'])
if args['--cv-timeout'] == "<forever>":
cv_timeout = None
else:
cv_timeout = float(args['--cv-timeout']) * 60 # Convert to seconds
scale_features = args['--scale-features']
verbose = args['--verbose']
run(params_config, features_path, observations, scoring, folds,
report, processes, cv_timeout, scale_features, verbose)
def main(argv=None):
args = docopt.docopt(__doc__, argv=argv)
logging.basicConfig(
level=logging.WARNING if not args['--debug'] else logging.DEBUG,
format='%(asctime)s %(levelname)s:%(name)s -- %(message)s'
)
features = yamlconf.import_module(args['<features>'])
session = mwapi.Session(args['--host'],
user_agent="Revscoring feature extractor utility")
if args['--login']:
sys.stderr.write("Log into " + args['--host'] + "\n")
sys.stderr.write("Username: "******"Password: "******"<stdin>":
rev_labels = read_rev_labels(sys.stdin)
else:
rev_labels = read_rev_labels(open(args['--rev-labels']))
if args['--value-labels'] == "<stdout>":
value_labels = sys.stdout
else:
value_labels = open(args['--value-labels'], 'w')
include_revid = bool(args['--include-revid'])
if args['--extractors'] == "<cpu count>":
extractors = cpu_count()
else:
extractors = int(args['--extractors'])
if args['--profile'] is not None:
profile_f = open(args['--profile'], 'w')
else:
profile_f = None
verbose = args['--verbose']
debug = args['--debug']
run(rev_labels, value_labels, features, extractor, include_revid,
extractors, profile_f, verbose, debug)
def _model_param_grid(params_config):
for name, config in params_config.items():
try:
Model = yamlconf.import_module(config['class'])
except Exception:
logger.warn("Could not load model {0}".format(config['class']))
logger.warn("Exception:\n" + traceback.format_exc())
continue
if not hasattr(Model, "train"):
logger.warn("Model {0} does not have a train() method.".format(
config['class']))
continue
param_grid = grid_search.ParameterGrid(config['params'])
yield name, Model, param_grid
def main(argv=None):
args = docopt.docopt(__doc__, argv=argv)
logging.basicConfig(
level=logging.INFO if not args['--debug'] else logging.DEBUG,
format='%(asctime)s %(levelname)s:%(name)s -- %(message)s')
params_config = yamlconf.load(open(args['<params-config>']))
features_path = args['<features>']
features = yamlconf.import_module(features_path)
label_decoder = util.DECODERS[args['--label-type']]
if args['--observations'] == "<stdin>":
observations_f = sys.stdin
else:
observations_f = open(args['--observations'])
observations = util.read_observations(observations_f, features,
label_decoder)
# Get a sepecialized scorer if we have one
scoring = metrics.SCORERS.get(args['--scoring'], args['--scoring'])
folds = int(args['--folds'])
if args['--report'] == "<stdout>":
report = sys.stdout
else:
report = open(args['--report'], "w")
if args['--processes'] == "<cpu-count>":
processes = multiprocessing.cpu_count()
else:
processes = int(args['--processes'])
if args['--cv-timeout'] == "<forever>":
cv_timeout = None
else:
cv_timeout = float(args['--cv-timeout']) * 60 # Convert to seconds
scale_features = args['--scale-features']
verbose = args['--verbose']
run(params_config, features_path, observations, scoring, folds, report,
processes, cv_timeout, scale_features, verbose)
def _model_param_grid(params_config):
for name, config in params_config.items():
try:
Model = yamlconf.import_module(config['class'])
except Exception:
logger.warn("Could not load model {0}"
.format(config['class']))
logger.warn("Exception:\n" + traceback.format_exc())
continue
if not hasattr(Model, "train"):
logger.warn("Model {0} does not have a train() method."
.format(config['class']))
continue
param_grid = grid_search.ParameterGrid(config['params'])
yield name, Model, param_grid
def _estimator_param_grid(params_config):
for name, config in params_config.items():
try:
EstimatorClass = yamlconf.import_module(config['class'])
estimator = EstimatorClass()
except Exception:
logger.warn("Could not load estimator {0}".format(config['class']))
logger.warn("Exception:\n" + traceback.format_exc())
continue
if not hasattr(estimator, "fit"):
logger.warn("Estimator {0} does not have a fit() method.".format(
config['class']))
continue
param_grid = grid_search.ParameterGrid(config['params'])
yield name, estimator, param_grid
def _estimator_param_grid(params_config):
for name, config in params_config.items():
try:
EstimatorClass = yamlconf.import_module(config['class'])
estimator = EstimatorClass()
except Exception:
logger.warn("Could not load estimator {0}"
.format(config['class']))
logger.warn("Exception:\n" + traceback.format_exc())
continue
if not hasattr(estimator, "fit"):
logger.warn("Estimator {0} does not have a fit() method."
.format(config['class']))
continue
param_grid = grid_search.ParameterGrid(config['params'])
yield name, estimator, param_grid
def from_config(self, config, name, section_key="languages"):
"""
Constructs a :class:`revscoring.languages.language.Language` from a
`dict`.
:Parameters:
config : dict
A configuration dictionary
name : str
The name of the sub-section in which to look for configuration
information
section_key : str
The top-level section key under which to look for `name`
"""
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
elif 'class' in section:
raise RuntimeError("Loading a language via class construction " + \
"not yet supported")
def from_config(self, config, name, section_key="languages"):
"""
Constructs a :class:`revscoring.languages.language.Language` from a
`dict`.
:Parameters:
config : dict
A configuration dictionary
name : str
The name of the sub-section in which to look for configuration
information
section_key : str
The top-level section key under which to look for `name`
"""
section = config[section_key][name]
if 'module' in section:
return yamlconf.import_module(section['module'])
elif 'class' in section:
raise RuntimeError("Loading a language via class construction " + \
"not yet supported")
def main(argv=None):
args = docopt.docopt(__doc__, argv=argv)
logging.basicConfig(
level=logging.WARNING if not args['--debug'] else logging.DEBUG,
format='%(asctime)s %(levelname)s:%(name)s -- %(message)s'
)
features = yamlconf.import_module(args['<features>'])
session = mwapi.Session(args['--host'],
user_agent="Revscoring feature extractor utility")
extractor = APIExtractor(session)
if args['--rev-labels'] == "<stdin>":
rev_labels = read_rev_labels(sys.stdin)
else:
rev_labels = read_rev_labels(open(args['--rev-labels']))
if args['--value-labels'] == "<stdout>":
value_labels = sys.stdout
else:
value_labels = open(args['--value-labels'], 'w')
include_revid = bool(args['--include-revid'])
if args['--extractors'] == "<cpu count>":
extractors = cpu_count()
else:
extractors = int(extractors)
verbose = args['--verbose']
debug = args['--debug']
run(rev_labels, value_labels, features, extractor, include_revid,
extractors, verbose, debug)
def from_config(cls, doc, name):
detector_class_path = doc['detectors'][name]['class']
Detector = yamlconf.import_module(detector_class_path)
return Detector.from_config(doc, name)
def from_config(cls, doc, name):
segmenter_class_path = doc['segmenters'][name]['class']
Segmenter = yamlconf.import_module(segmenter_class_path)
return Segmenter.from_config(doc, name)
def from_config(cls, doc, name):
segmenter_class_path = doc['segmenters'][name]['class']
Segmenter = yamlconf.import_module(segmenter_class_path)
return Segmenter.from_config(doc, name)
def from_config(cls, doc, name):
tokenizer_class_path = doc['tokenizers'][name]['class']
Tokenizer = yamlconf.import_module(tokenizer_class_path)
return Tokenizer.from_config(doc, name)
def from_config(cls, doc, name):
detector_class_path = doc['detectors'][name]['class']
Detector = yamlconf.import_module(detector_class_path)
return Detector.from_config(doc, name)
def from_config(cls, doc, name):
tokenizer_class_path = doc['tokenizers'][name]['class']
Tokenizer = yamlconf.import_module(tokenizer_class_path)
return Tokenizer.from_config(doc, name)