def buildAllCorpus(element_type='photos', time_interval_length=14, debug=False, paras={}):
# return a dict = {region : its local corpus}
assert element_type in ['photos', 'tweets']
all_corpus = {}
if element_type == 'photos':
config = InstagramConfig
else:
config = TwitterConfig
coordinates = [config.min_lat, config.min_lng,
config.max_lat, config.max_lng]
nyc = Region(coordinates)
region_list = nyc.divideRegions(25, 25)
region_list = nyc.filterRegions(region_list, test=True, n=25, m=25, element_type=element_type)
# 14 days ago
now = int(tool.getCurrentStampUTC())
num = 0
for region in region_list:
if debug and num > 0:
# speed up the debugging
pass
else:
cor = Corpus()
cor.buildCorpus(region, [now - time_interval_length * 3600 * 24, now], element_type, paras)
all_corpus[region.getKey()] = cor
num += 1
print 'build corpus %d' % (num)
return all_corpus
def test():
coordinates = [InstagramConfig.photo_min_lat,
InstagramConfig.photo_min_lng,
InstagramConfig.photo_max_lat,
InstagramConfig.photo_max_lng
]
huge_region = Region(coordinates)
alarm_region_size = 25
regions = huge_region.divideRegions(25, 25)
filtered_regions = huge_region.filterRegions(region_list=regions, test=True, n=alarm_region_size,
m=alarm_region_size)
for i in range(1):
test_region = regions[i]
test_region._region['min_lat'] = 40.7329
test_region._region['min_lng'] = -73.9957
test_region._region['max_lat'] = 40.7383
test_region._region['max_lng'] = -73.9844
test_region.display()
ts = TwitterTimeSeries(test_region, '1364829908', '1365693908')
ts = ts.buildTimeSeries()
for d in ts:
print d
