def __init__(self, recreate=False):
datamodel = Datamodel()
self.table, self.table_config = datamodel.population()
self.postgres = Postgresql(user_name='postgres',
password='******',
host='localhost',
port='5432',
db='TestProject')
self.postgres.initialize_table(self.table,
recreate=False,
**self.table_config)
self.googlegeo = GoogleGeo()
# myan: only get major cities data once per request
self.major_cities_postgres = Postgresql(user_name='postgres',
password='******',
host='localhost',
port='5432',
db='TestProject')
self.major_cities = self.major_cities_postgres.get(
"select * from TestMajorCities")
self.all_states = self.major_cities['state'].values
self.all_cities = self.major_cities['city'].values
self.all_lats = self.major_cities['lat'].values
self.all_lngs = self.major_cities['lng'].values
self.all_population = self.major_cities['population'].values
def push_major_cities():
all_cities = requests.get(
'http://api.census.gov/data/2013/acs3?get=NAME,B01001_001E&for=place:*'
)
major_cities, _ = _parse_city_population(all_cities.json())
major_cities = major_cities.loc[
major_cities.population >=
100000] # myan: major city should have more than 100000
major_cities = _get_city_info(major_cities)
postgres = Postgresql(user_name='postgres',
password='******',
host='localhost',
port='5432',
db='TestProject')
field_types = {
'place_id': 'TEXT',
'city': 'TEXT',
'state': 'TEXT',
'population': 'INT',
'lat': 'FLOAT',
'lng': 'FLOAT'
}
postgres.initialize_table('TestMajorCities',
fields_types=field_types,
primary_key='place_id',
not_null_fields=['place_id', 'city', 'state'],
recreate=False)
postgres.put_dataframe(major_cities, field_types)
def test_parse_list_values(self):
pg = Postgresql()
insert = pg.parse_values_list(
[{
'gsId': 100,
'name': 'Max',
'zip_code': 123,
'state': 'CA',
'gsRating': 3.5
}, {
'gsId': 101,
'name': 'Tez',
'zip_code': 123,
'state': 'CA',
'gsRating': 8.5
}], {
'gsId': 'INT',
'zip_code': 'INT',
'state': 'TEXT',
'name': 'TEXT',
'gsRating': 'FLOAT'
},
field_list=['gsId', 'name', 'gsRating'])
assert insert == "(100, 'Max', 3.5),(101, 'Tez', 8.5)"
def __init__(self, key=None):
if key is None:
self.api_key = _get_great_schools_api_key()
else:
self.api_key = key
# myan: initialize postgresql
datamodel = Datamodel()
self.table, self.table_config = datamodel.great_schools()
self.postgres = Postgresql(user_name='postgres',
password='******',
host='localhost',
port='5432',
db='TestProject')
self.postgres.initialize_table(self.table, recreate=False, **self.table_config)
def test_make_schema_string(self):
pg = Postgresql()
fields = {
'gsid': 'INT',
'zip_code': 'INT',
'state': 'TEXT',
'name': 'TEXT',
'gsrating': 'FLOAT'
}
primary_key = 'gsid'
not_null_fields = ['gsid', 'zip_code', 'state', 'name']
schema_string = pg.make_schema_string(fields,
primary_key=primary_key,
not_null_fields=not_null_fields)
assert schema_string == '(gsid INT PRIMARY KEY NOT NULL,zip_code INT NOT NULL,state TEXT NOT NULL,name TEXT NOT NULL,gsrating FLOAT)'
class ZillowDataHandler:
def __init__(self, table=None, recreate=False):
datamodel = Datamodel()
if table is None:
table, table_config = datamodel.zipcode_timeseries()
else:
_, table_config = datamodel.zipcode_timeseries()
self.time_series_postgres = self._initialize_postgres(
(table, table_config), recreate=recreate)
def _initialize_postgres(self, (table, config), recreate=False):
postgres = Postgresql(user_name='postgres',
password='******',
host='localhost',
port='5432',
db='TestProject')
postgres.initialize_table(table, recreate=recreate, **config)
return postgres
def get_data_from_zillow(self):
# TODO: complete downloading data from web part
pass
def _drop_columns(self,
data,
all_columns=('RegionID', 'City', 'Metro', 'SizeRank')):
to_drop = []
for column in all_columns:
if column in data.columns:
class Plotter:
def __init__(self,
min_price=150000,
max_price=300000,
top_percentage=0.25,
top_max_num_entries=30):
self.MIN_PRICE = min_price
self.MAX_PRICE = max_price
self.TOP_PERCENTAGE = top_percentage
self.TOP_MAX_NUM_ENTRIES = top_max_num_entries
datamodel = Datamodel()
self.time_series_postgres = self._initialize_postgres(
datamodel.zipcode_timeseries())
self.population_postgres = self._initialize_postgres(
datamodel.population())
self.great_schools_postgres = self._initialize_postgres(
datamodel.great_schools())
self.zipcode_timeseries = None
self.top_zipcodes_timeseries = None
self.rest_zipcodes_timeseries = None
self.top_zipcodes_school_data = None
self.top_zipcodes_population_data = None
self.all_months_timeseries = None
def reset(self):
self.zipcode_timeseries = None
self.top_zipcodes_timeseries = None
self.rest_zipcodes_timeseries = None
self.top_zipcodes_school_data = None
self.top_zipcodes_population_data = None
self.all_months_timeseries = None
def _get_time_series_data(self, year_month=201510):
suitable_states = self.time_series_postgres.get(
"select * from {table} where year_month={year_month} and median_price < {max_price} and median_price > {min_price}"
.format(table=self.time_series_postgres.table,
year_month=year_month,
max_price=self.MAX_PRICE,
min_price=self.MIN_PRICE))
suitable_states[
'gross_yield_pct'] = suitable_states.median_rent * 12 / suitable_states.median_price * 100
suitable_states = suitable_states.sort('gross_yield_pct',
ascending=False)
top_index = min(int(len(suitable_states) * self.TOP_PERCENTAGE),
self.TOP_MAX_NUM_ENTRIES)
self.zipcode_timeseries = suitable_states
self.top_zipcodes_timeseries = suitable_states[:top_index]
self.rest_zipcodes_timeseries = suitable_states[top_index:]
def _get_population_data(self):
if self.top_zipcodes_timeseries is None:
self._get_time_series_data()
p = Population(recreate=False)
self.top_zipcodes_population_data = p.run(
addresses=self.top_zipcodes_timeseries.zip_code.values)
def _get_top_zipcodes_school_data(self):
all_states = self.top_zipcodes_timeseries.state.values
all_zipcodes = self.top_zipcodes_timeseries.zip_code.values
unique_states = self.top_zipcodes_timeseries.state.unique()
requests = []
for state in unique_states:
unique_zipcodes = np.unique(all_zipcodes[all_states == state])
for zip_code in unique_zipcodes:
requests.append(
dict(db_configs=dict(
postgres=self.great_schools_postgres,
query=
"select * from {table} where state='{state}' and zip_code={zip_code};"
.format(table=self.great_schools_postgres.table,
state=state,
zip_code=zip_code)),
api_configs=dict(api=GreatSchools,
api_key=None,
api_args=dict(state=state,
zip_code=zip_code,
limit=20))))
mission_control = MissionControl()
gs_data = mission_control.request_data(user_requests=requests)
gs_data_df = pd.DataFrame()
for entry in gs_data:
gs_data_df = gs_data_df.append(entry)
self.top_zipcodes_school_data = pd.DataFrame(
gs_data_df.groupby('zip_code')['gsrating'].mean())
self.top_zipcodes_school_data[
'zip_code'] = self.top_zipcodes_school_data.index
def _initialize_postgres(self, (table, config)):
postgres = Postgresql(user_name='postgres',
password='******',
host='localhost',
port='5432',
db='TestProject')
postgres.initialize_table(table, recreate=False, **config)
return postgres
