def set_up(self):
data_df = read_csv("datapoints/tests/_data/calc_data.csv")
self.create_metadata()
self.user = User.objects.get(username="******")
self.test_df = data_df[data_df["is_raw"] == 1]
self.target_df = data_df[data_df["is_raw"] == 0]
ltr = LocationTreeCache()
ltr.main()
def set_up(self):
data_df = read_csv('datapoints/tests/_data/calc_data.csv')
self.create_metadata()
self.user = User.objects.get(username="******")
self.test_df = data_df[data_df['is_raw'] == 1]
self.target_df = data_df[data_df['is_raw'] == 0]
ltr = LocationTreeCache()
ltr.main()
def test_cache_location_tree(self):
'''
The point of this test is that it the cache_location_tree function
creates database rows that represents each parent / child relationship
between all locations, from the stored information.
That is, if we say that NYC is a sub-location of NYState, and NYState
is a sub-location of U.S.A. , then that means that we should have a row
in the LocationTree table rebresenting the fact that both Nystate AND
U.S.A. are parent_location's on New York City.
'''
location_batch = []
office = Office.objects.create(name='not important')
location_type_country = LocationType.objects.create(name='Country',\
admin_level=0,id=1)
location_type_state = LocationType.objects.create(name='State',\
admin_level=1,id=2)
location_type_city = LocationType.objects.create(name='City',\
admin_level=2,id=3)
location_type_city = LocationType.objects.create(name='Neighborhood',\
admin_level=3,id=4)
location_data = {
'id':[1,2,3,4,5,6,7,8],
'location_type_id':[1,2,3,2,2,2,3,4],
'name': ['U.S.A.','New York State','New York City','Texas',\
'California','Maine','Portland, ME','Brooklyn'],
'parent_location_id': [None,1,2,1,1,1,6,3]
}
df = DataFrame.from_dict(location_data)
df.set_index('id', inplace=True, drop=False, append=False)
no_nan_df = df.where((notnull(df)), None)
for ix, loc in no_nan_df.iterrows():
location_batch.append(Location(**{
'id':loc.id,
'name':loc.name,
'location_code': unicode(loc.name).replace(' ',''),
'office_id':office.id,
'location_type_id':loc.location_type_id,
'parent_location_id':loc.parent_location_id
}))
Location.objects.bulk_create(location_batch)
## now use the following function to transform the location_tree ##
ltc = LocationTreeCache()
ltc.main()
location_tree_in_db = LocationTree.objects.all()\
.values_list('location_id','parent_location_id')
## the ultimate parent, should have a with itself is parent ##
self.assertTrue((1,1) in location_tree_in_db)
## Brooklyn However should not be the parent of itself ##
self.assertFalse((8,8) in location_tree_in_db)
## new york city has a parent of NY State AND U.S.A ##
self.assertTrue((3,1) in location_tree_in_db)
self.assertTrue((3,2) in location_tree_in_db)
## Portland ME has a parent of Maine AND U.S.A ##
self.assertTrue((7,1) in location_tree_in_db)
self.assertTrue((7,6) in location_tree_in_db)
# ## new york state and maine are children of USA #
self.assertTrue((2,1) in location_tree_in_db)
self.assertTrue((6,1) in location_tree_in_db)
## Brooklkyn has, NYC, NYState and USA as parents ##
self.assertTrue((8,3) in location_tree_in_db)
self.assertTrue((8,2) in location_tree_in_db)
self.assertTrue((8,1) in location_tree_in_db)
def test_cache_location_tree(self):
'''
The point of this test is that it the cache_location_tree function
creates database rows that represents each parent / child relationship
between all locations, from the stored information.
That is, if we say that NYC is a sub-location of NYState, and NYState
is a sub-location of U.S.A. , then that means that we should have a row
in the LocationTree table rebresenting the fact that both Nystate AND
U.S.A. are parent_location's on New York City.
'''
location_batch = []
office = Office.objects.create(name='not important')
location_type_country = LocationType.objects.create(name='Country',\
admin_level=0,id=1)
location_type_state = LocationType.objects.create(name='State',\
admin_level=1,id=2)
location_type_city = LocationType.objects.create(name='City',\
admin_level=2,id=3)
location_type_city = LocationType.objects.create(name='Neighborhood',\
admin_level=3,id=4)
location_data = {
'id':[1,2,3,4,5,6,7,8],
'location_type_id':[1,2,3,2,2,2,3,4],
'name': ['U.S.A.','New York State','New York City','Texas',\
'California','Maine','Portland, ME','Brooklyn'],
'parent_location_id': [None,1,2,1,1,1,6,3]
}
df = DataFrame.from_dict(location_data)
df.set_index('id', inplace=True, drop=False, append=False)
no_nan_df = df.where((notnull(df)), None)
for ix, loc in no_nan_df.iterrows():
location_batch.append(
Location(
**{
'id': loc.id,
'name': loc.name,
'location_code': unicode(loc.name).replace(' ', ''),
'office_id': office.id,
'location_type_id': loc.location_type_id,
'parent_location_id': loc.parent_location_id
}))
Location.objects.bulk_create(location_batch)
## now use the following function to transform the location_tree ##
ltc = LocationTreeCache()
ltc.main()
location_tree_in_db = LocationTree.objects.all()\
.values_list('location_id','parent_location_id')
## the ultimate parent, should have a with itself is parent ##
self.assertTrue((1, 1) in location_tree_in_db)
## Brooklyn However should not be the parent of itself ##
self.assertFalse((8, 8) in location_tree_in_db)
## new york city has a parent of NY State AND U.S.A ##
self.assertTrue((3, 1) in location_tree_in_db)
self.assertTrue((3, 2) in location_tree_in_db)
## Portland ME has a parent of Maine AND U.S.A ##
self.assertTrue((7, 1) in location_tree_in_db)
self.assertTrue((7, 6) in location_tree_in_db)
# ## new york state and maine are children of USA #
self.assertTrue((2, 1) in location_tree_in_db)
self.assertTrue((6, 1) in location_tree_in_db)
## Brooklkyn has, NYC, NYState and USA as parents ##
self.assertTrue((8, 3) in location_tree_in_db)
self.assertTrue((8, 2) in location_tree_in_db)
self.assertTrue((8, 1) in location_tree_in_db)