def time_seq_count():
time_seqs_per_label = {}
for fname in os.listdir(data_dir):
full_fname = os.path.join(data_dir, fname)
for line in open(full_fname):
created_at, user_id, label, text = line.rstrip('\n').split('\t')
time_seq = timeSeq(created_at)
if not ok_without_rt(text):
continue
if label not in time_seqs_per_label:
time_seqs_per_label[label] = []
time_seqs_per_label[label].append(time_seq)
# Let's aggregate into a single table
labels = map(itemgetter(0), time_seqs_per_label.items())
#print labels
counts_per_time_unit = {}
for label, time_seqs in time_seqs_per_label.iteritems():
c = sorted(Counter(time_seqs).most_common(), key=itemgetter(0))
for (time_seq, count) in c:
if time_seq not in counts_per_time_unit:
counts_per_time_unit[time_seq] = {}
for l in labels:
counts_per_time_unit[time_seq][l] = 0
counts_per_time_unit[time_seq][label] = count
return (counts_per_time_unit, labels)
def extract_specific_time_music(week_num, music_label):
user_ids = []
for fname in os.listdir(data_dir):
full_fname = os.path.join(data_dir, fname)
for line in open(full_fname):
created_at, user_id, label, text = line.rstrip('\n').split('\t')
time_seq = timeSeq(created_at)
if time_seq == week_num and label == music_label:
user_ids.append(user_id)
user_ids = list(set(user_ids))
return user_ids
def extract_show(show_label):
user_ids = []
for fname in os.listdir(data_dir):
full_fname = os.path.join(data_dir, fname)
for line in open(full_fname):
created_at, user_id, label, text = line.rstrip('\n').split('\t')
if ok_without_rt(text):
time_seq = timeSeq(created_at)
if label == show_label:
user_ids.append(user_id)
user_ids = list(set(user_ids))
return user_ids
