def main():
optparser = OptionParser(usage="""
%prog [OPTIONS] TRAIN_FILE TEST_FILE""")
opts, args = optparser.parse_args()
for session in SessionReader().open(args[0]):
print session.user_id
def main():
optparser = OptionParser(usage="""
%prog [OPTIONS] TRAIN_FILE STATISTIC_FILE""")
optparser.add_option('-d',
'--description',
dest='description_file',
metavar='FILE',
type='string',
default=None,
help='filepath for features description')
opts, args = optparser.parse_args()
urls = {}
for line in open(args[1]):
s = line.strip().split('\t')
url = int(s[0])
ctr = float(s[1])
shows = int(s[2])
urls[url] = (ctr, shows)
url_features_calcer = UrlFeatureCalcer(urls)
for session in SessionReader().open(args[0]):
print "\t".join(map(str, url_features_calcer.calc_features(session)))
if opts.description_file is not None:
output_file = open(opts.description_file, 'w')
print >> output_file, "\n".join(
map(str, url_features_calcer.get_description()))
output_file.close()
def main():
optparser = OptionParser(usage="""
%prog [OPTIONS] DATA_FILE""")
opts, args = optparser.parse_args()
data_file = args[0]
queries_count = {}
uniq_switch_count = {}
switch_count = {}
for session in SessionReader().open(data_file):
for query in session.queries:
if query.query_id not in queries_count:
queries_count[query.query_id] = 0
uniq_switch_count[query.query_id] = 0
switch_count[query.query_id] = 0
queries_count[query.query_id] += 1
uniq_switch_count[query.query_id] += int(session.has_switch())
switch_count[query.query_id] += len(session.switches)
for query_id in queries_count.keys():
print "\t".join(
map(str, [
query_id, queries_count[query_id], uniq_switch_count[query_id],
switch_count[query_id]
]))
def main():
users_info = {}
for session in SessionReader().open(sys.argv[1]):
user_id = session.user_id
if user_id not in users_info:
users_info[user_id] = UserInfo(user_id)
users_info[user_id].add(session)
for user, user_info in users_info.iteritems():
user_info.flush()
def main():
optparser = OptionParser(usage="""
%prog [OPTIONS] TRAIN_FILE TEST_FILE""")
optparser.add_option('-d',
'--description',
dest='description_file',
metavar='FILE',
type='string',
default=None,
help='filepath for features description')
opts, args = optparser.parse_args()
session_calcer = SessionFeatureCalcer()
for session in SessionReader().open(args[0]):
print "\t".join(map(str, session_calcer.calc_features(session)))
if opts.description_file is not None:
output_file = open(opts.description_file, 'w')
print >> output_file, "\n".join(
map(str, session_calcer.get_description()))
output_file.close()
def main():
optparser = OptionParser(usage="""
%prog [OPTIONS] DATA_FILE TOP_SIZE RESULT_SIZE COEFFICIENT""")
opts, args = optparser.parse_args()
data_file = args[0]
top_size = int(args[1])
result_size = int(args[2])
coefficient = None
if len(args) == 4:
coefficient = float(args[3])
url_in_top = {}
url_clicks = {}
for session in SessionReader().open(data_file):
for query in session.queries:
for url in query.urls[:top_size]:
if url not in url_in_top:
url_in_top[url] = 0
url_clicks[url] = 0
url_in_top[url] += 1
for click in query.clicks:
if click.url_id == url:
url_clicks[url] += 1
break
urls = []
for url in url_clicks:
urls.append(
(url, 1.0 * url_clicks[url] / url_in_top[url], url_in_top[url]))
urls.sort(key=lambda x: (x[1], x[2]))
if coefficient is not None:
urls.sort(
key=lambda x: (coefficient * x[1] + (1 - coefficient) * x[2]))
for url_info in urls[-result_size:]:
print "\t".join(map(str, url_info))