class testLocationsKDTree(unittest.TestCase):
def setUp(self):
self.test = LocationsKDTree()
# The 15 points manually selected from the San Francisco Map
self.test.mercatorPointsArray = [(552084.9682165304, 4182494.1939107366), (551922.8858612546, 4182574.0555520994), (551687.9949461814, 4182217.968801949),
(552775.8082342255, 4183336.7417580322), (553222.9632670761, 4183422.393102693), (552198.7222675259, 4184064.8048690264),
(549362.9961397024, 4181902.574203599), (548023.1142878283, 4180961.314933798), (551183.5342344284, 4184494.9211385823),
(548547.6683833781, 4179180.1576283425), (549558.9617763544, 4184499.9538624967), (553803.7182814779, 4180131.3817638056),
(552354.47799122, 4184532.357535243), (550631.7088304324, 4182068.2753886213), (550951.7531905753, 4181648.8476454713)]
self.test.build_kd_tree()
def test_lookup_location(self):
"""
Search for the 5 points closest to the chosen point on the San Francisco Map
Chosen point = (552048.7215749588, 4182698.120068086)
The known/expected 5 closest points:
(551922.8858612546, 4182574.0555520994)
(552084.9682165304, 4182494.1939107366)
(551687.9949461814, 4182217.968801949)
(552775.8082342255, 4183336.7417580322)
(552198.7222675259, 4184064.8048690264)
"""
nearest_nn_points = self.test.find_nn((552048.7215749588, 4182698.120068086),5)
self.assertEqual(nearest_nn_points[0],(551922.8858612546, 4182574.0555520994))
self.assertEqual(nearest_nn_points[1],(552084.9682165304, 4182494.1939107366))
self.assertEqual(nearest_nn_points[2],(551687.9949461814, 4182217.968801949))
self.assertEqual(nearest_nn_points[3],(552775.8082342255, 4183336.7417580322))
self.assertEqual(nearest_nn_points[4],(552198.7222675259, 4184064.8048690264))
def setUp(self):
self.app = create_app('testing')
self.app_context = self.app.app_context()
self.app_context.push()
self.client = self.app.test_client()
# An inverted index, mapping UTM (Universal Transverse Mercator) lat/lng
# values geolocating each film, to a LIST of films that were filmed
# at that location.
#
# IMPORTANT!: This is a shared data-structure, built only at startup,
# that is READ-ONLY by all, and so can be safely shared.
invertedPointLocationIndex = InvertedPointLocationIndex()
invertedPointLocationIndex.build_inverted_location_index()
# A KD tree, implemented using the scipy package's kdtree implementation
# under the hood, to allow for fast O(ln) queries of 2D point data. The
# points that it stores are geocoded locations coded in UTM to allow them
# to be treated as 2D points to an approximation.
#
# IMPORTANT!: This is a shared data-structure, built only at startup,
# that is READ-ONLY by all, and so can be safely shared.
filmLocationsKDTree = LocationsKDTree()
filmLocationsKDTree.load_point_data()
filmLocationsKDTree.build_kd_tree()
self.app_context.g.filmlocationsKDTree = filmLocationsKDTree
self.app_context.g.invertedPointLocationIndex = invertedPointLocationIndex
def setUp(self):
self.test = LocationsKDTree()
# The 15 points manually selected from the San Francisco Map
self.test.mercatorPointsArray = [(552084.9682165304, 4182494.1939107366), (551922.8858612546, 4182574.0555520994), (551687.9949461814, 4182217.968801949),
(552775.8082342255, 4183336.7417580322), (553222.9632670761, 4183422.393102693), (552198.7222675259, 4184064.8048690264),
(549362.9961397024, 4181902.574203599), (548023.1142878283, 4180961.314933798), (551183.5342344284, 4184494.9211385823),
(548547.6683833781, 4179180.1576283425), (549558.9617763544, 4184499.9538624967), (553803.7182814779, 4180131.3817638056),
(552354.47799122, 4184532.357535243), (550631.7088304324, 4182068.2753886213), (550951.7531905753, 4181648.8476454713)]
self.test.build_kd_tree()
# values geolocating each film, to a LIST of films that were filmed
# at that location.
#
# IMPORTANT!: This is a shared data-structure, built only at startup,
# that is READ-ONLY by all, and so can be safely shared.
invertedPointLocationIndex = InvertedPointLocationIndex()
invertedPointLocationIndex.build_inverted_location_index()
# A KD tree, implemented using the scipy package's kdtree implementation
# under the hood, to allow for fast O(ln) queries of 2D point data. The
# points that it stores are geocoded locations coded in UTM to allow them
# to be treated as 2D points to an approximation.
#
# IMPORTANT!: This is a shared data-structure, built only at startup,
# that is READ-ONLY by all, and so can be safely shared.
filmLocationsKDTree = LocationsKDTree()
filmLocationsKDTree.load_point_data()
filmLocationsKDTree.build_kd_tree()
# If the request is for the endpoint 'films_near_me,' which is seeking
# the 7 films closest to a user's location, only then do we bother to
# load the request with the globally existing KD tree and inverted
# index.
@app.before_request
def before_request():
print "before request ..."
if request.endpoint == 'films_near_me':
g.filmlocationsKDTree = filmLocationsKDTree
g.invertedPointLocationIndex = invertedPointLocationIndex
def make_shell_context():
