tweets.py

import csv
import collections
import datetime

dates = {}

with open('tweets.csv') as fh:
    fh.readline()
    reader = csv.reader(fh)
    for row in reader:
         date,time,zone = row[3].split(' ')
         if date < "2012":
             continue
         y,m,d = date.split('-')
         date = datetime.datetime(int(y),int(m), int(d))
         if date not in dates:
             dates[date] = collections.Counter()
         hour = int(time[:2])
         dates[date][hour]+=1
             
oldest,newest = min(dates.keys()), max(dates.keys())

for d in range((newest-oldest).days):
    day = oldest + datetime.timedelta(days=d)
    if day not in dates:
        dates[day] = {}

for date in sorted(dates):
    day = ["_"] * 48
    for hour, count in dates[date].iteritems():
        day[hour] = "*"
    next = date + datetime.timedelta(days=1)
    if next in dates:
        for hour, count in dates[next].iteritems():
            day[hour+24] = "*"
    print('"{:04}-{:02}-{:02}",{}'.format(date.year,date.month,date.day, ",".join(day)))