p.add_argument("--verbose",
action="store_true",
dest="verbose",
help="Verbose mode [Off]")
opts = p.parse_args()
# Parameters
out = sys.stdout
if opts.output:
try:
out = open(opts.output)
except:
p.error('Output parameter could not been open: {0}'\
.format(opts.output))
with open(opts.conf, 'r') as stream:
conf = yaml.load(stream)
final = {}
for answersfile in opts.Answers:
key_word = basename(answersfile)
ans = docread.loadanswers(answersfile)
for k, v in ans.iteritems():
try:
final[k] += v * conf[key_word][k[:2]]
except KeyError:
final[k] = v * conf[key_word][k[:2]]
for k, v in final.iteritems():
print k, "{0:0.3f}".format(v)
help="Output [STDOUT]")
p.add_argument( "--verbose",
action="store_true", dest="verbose",
help="Verbose mode [Off]")
opts = p.parse_args()
# Parameters
out = sys.stdout
if opts.output:
try:
out = open(opts.output)
except:
p.error('Output parameter could not been open: {0}'\
.format(opts.output))
gs_ = docread.loadanswers(opts.GS,code=codes[opts.language][opts.genre])
sys = docread.loadanswers(opts.SYS,code=codes[opts.language][opts.genre])
gs={}
for k in sys.keys():
gs[k]=gs_[k]
#probas = docread.loadproba(args[1])
tp=0
fp=0
fn=0
total=0
yeas=0
noes=0
sin_contestar=0
tpd={}
# Loading main files -------------------------------------------------
# load problems or problem
verbose('Loading files')
problems=docread.problems(
docread.dirproblems(dirname,known_pattern,unknown_pattern,_ignore,
code=codes[opts.language][opts.genre]))
# Loading answers file only for DEVELOPMENT OR TRAINNING MODE
if opts.mode.startswith("train") or opts.mode.startswith("devel"):
if opts.Answers:
answers_file=opts.Answers
else:
answers_file="{0}/{1}".format(dirname,opts.answers)
verbose('Loading answer file: {0}'.format(answers_file))
answers = docread.loadanswers(answers_file,_ignore,
code=codes[opts.language][opts.genre])
# Checking for consistency
if not len(problems) == len(answers):
p.error("Not match for number of problems({0}) and \
answers({1})".format(len(problems),len(answers)))
samples=[]
classes=[]
# Create the representations for known and unknown documents
for id,(ks,uks) in problems:
verbose('Reading from : {0}'.format(id))
# Load unknown
if len(uks) > 1:
p.error("More than one unknown file for {0}".format(id))
verbose('Loading files to ignore frm: .ignore')
with open('.ignore') as file:
_ignore=file.read().readlines()
# load problems or problem
problems=docread.dirproblems(dirname,opts.known,opts.unknown,_ignore)
# TRAINNING MODE
if opts.mode.startswith("train"):
# Loading answers file
if not len(args)==2:
p.error("Answers needed for train mode")
verbose('Loading answer file: {0}'.format(args[1]))
answers = docread.loadanswers(args[1])
# Checking for consistency
if not len(problems) == len(answers):
p.error("Not match for number of problems({0}) and \
answers({1})".format(len(problems),len(answers)))
# Loading weights or initializing
if opts.weights:
verbose('Loading weights file: {0}'.format(args[1]))
# TODO load weights when given file
else:
WS=W.Weights()
samples=[]
opts = p.parse_args()
# Parameters
out = sys.stdout
if opts.output:
try:
out = open(opts.output)
except:
p.error('Output parameter could not been open: {0}'\
.format(opts.output))
with open(opts.conf, 'r') as stream:
conf=yaml.load(stream)
final={}
for answersfile in opts.Answers:
key_word=basename(answersfile)
ans = docread.loadanswers(answersfile)
for k,v in ans.iteritems():
try:
final[k]+=v*conf[key_word][k[:2]]
except KeyError:
final[k]=v*conf[key_word][k[:2]]
for k,v in final.iteritems():
print k, "{0:0.3f}".format(v)
# Loading ignore if exists
_ignore=[]
if os.path.exists('.ignore'):
verbose('Loading files to ignore from: .ignore')
with open('.ignore') as file:
for line in file:
_ignore.append(line.strip())
# Loading stopwords if exits
stopwords=[]
if os.path.exists(opts.stopwords):
verbose('Loading stopwords: ',opts.stopwords)
stopwords=docread.readstopwords(opts.stopwords)
else:
info('Stopwords file not found assuming, emtpy',opts.stopwords)
# Loading main files -------------------------------------------------
# load problems or problem
verbose('Loading files')
problems=docread.problems(
docread.dirproblems(dirname,known_pattern,unknown_pattern,_ignore,
code=codes[opts.language][opts.genre]))
gs = docread.loadanswers(opts.GS,code=codes[opts.language][opts.genre])
sys = docread.loadanswers(opts.SYS,code=codes[opts.language][opts.genre])
console=AuthorIdCLI()
console.cmdloop()
