pastebin - collaborative debugging

#!/usr/bin/python

'''The following script cleans up data from the yfd tab-delimited file in a
cursory way and creates a numpy array of duration, start, and end time.
'''

import numpy as np
import csv
import pytz
from pytz import timezone
from datetime import datetime, timedelta

file = open('ethandataOct28.txt')
file.readline()

#create numpy array of data
types = {'names':('name','value','unit','time'),
         'formats':('S10', 'S5', 'S5', 'S20')}

ethan = np.loadtxt(file, dtype=types, delimiter='\t')

sleep_mask = np.logical_or(ethan['name']=='gnight', ethan['name']=='gmorning')
sleep_index = np.where(sleep_mask)[0]
sleep_raw = ethan[sleep_mask]

#find potential errors in data (i.e., there are 2 gnight's in a row)
errors_index = np.where(sleep_raw['name'][:-1] == sleep_raw['name'][1:])
errors = sleep_raw[errors_index]

if errors.size != 0:
    print "There are duplicate gnight/gmorning entries. Check the data!"

#create array of cleaned data
sleep_clean = np.column_stack((sleep_raw['name'], sleep_raw['time']))

time_stringlist = sleep_clean[:,1].tolist()

time_objectlist=[]
for item in time_stringlist:
    time_objectlist.append(datetime.strptime(item, "%Y-%m-%d %H:%M:%S"))

time_objectarray = np.array(time_objectlist)

sleep_events = np.column_stack((sleep_clean[:,0], time_objectarray))

#a little logic to make sure I calculate time asleep and not time awake
if sleep_events[0][0] == 'gmorning':
    gmorning_objects = sleep_events[::2,1]
    gnight_objects = sleep_events[1::2,1]
else:
    gmorning_objects = sleep_events[1::2,1]
    gnight_objects = sleep_events[2::2,1]

sleep_durations = gmorning_objects - gnight_objects
duration_table = np.column_stack((sleep_durations,
                                  gnight_objects, gmorning_objects))