def run(timezones_amount=20):
# dataframe takes list of dicts
frame = pd.DataFrame(load_jsoned_file('bitlygov_ua.json'))
print frame
tz_counts = frame['tz'].value_counts().fillna('Missing')
pprint(tz_counts[:timezones_amount])
print 'ploting...'
plot_top(tz_counts, timezones_amount)
def run():
frame = pd.DataFrame(misc.load_jsoned_file('bitlygov_ua.json'))
# get not null fields from 'a' field
cframe = frame[frame.a.notnull()]
# np.where first is the condition (has Windows in string)
# Group to Windiws, Non Windows fields
# example
# In [42]: np.where([1,0,1,1], "tak", "nie")
# Out[42]:
# array(['tak', 'nie', 'tak', 'tak'],
# dtype='|S3')
oss = np.where(cframe['a'].str.contains('Windows'), 'Windows', 'Non Windows')
# grouping timezones by oss
tz_by_os = cframe.groupby([cframe['tz'], oss])
# size.unstack ????
agg_tzs = tz_by_os.size().unstack().fillna(0)
# ????
indexer = agg_tzs.sum(1).argsort()
# ????
count_subset = agg_tzs.take(indexer)[-10:]
count_subset.plot(kind='barh', stacked=True)
