def test_random_state():
# no seeds
reno = Community.from_census(msa_fips="39900", datastore=DataStore())
r1 = reno.cluster(columns=columns, method="kmeans", n_clusters=5)
r2 = reno.cluster(columns=columns, method="kmeans", n_clusters=5)
card1 = r1.gdf.groupby("kmeans").count()["geoid"].values
card1.sort()
card2 = r2.gdf.groupby("kmeans").count()["geoid"].values
card2.sort()
# test that the cardinalities are different
np.testing.assert_raises(
AssertionError, np.testing.assert_array_equal, card1, card2
)
# seeds
reno = Community.from_census(msa_fips="39900", datastore=DataStore())
seed = 10
r1 = reno.cluster(columns=columns, method="kmeans", n_clusters=5, random_state=seed)
r2 = reno.cluster(columns=columns, method="kmeans", n_clusters=5, random_state=seed)
card1 = r1.gdf.groupby("kmeans").count()["geoid"].values
card1.sort()
card2 = r2.gdf.groupby("kmeans").count()["geoid"].values
card2.sort()
# test that the cardinalities are identical
np.testing.assert_array_equal(card1, card2)
def test_harmonize_area():
la = Community.from_census(county_fips="06037")
harmonized = la.harmonize(2000,
extensive_variables=["n_total_housing_units"],
intensive_variables=["p_vacant_housing_units"],
raster=local_raster)
assert_allclose(
harmonized.gdf[harmonized.gdf.year ==
2000].n_total_housing_units.sum(),
3271578.974605,
atol=600,
)
assert_allclose(
harmonized.gdf[harmonized.gdf.year ==
1990].n_total_housing_units.sum(),
3163560.996240,
)
assert_allclose(
harmonized.gdf[harmonized.gdf.year ==
2010].n_total_housing_units.sum(),
3441415.997327,
)
assert_allclose(harmonized.gdf.p_vacant_housing_units.sum(),
33011.58879,
rtol=1e-03)
def test_harmonize_area_weighted():
balt = Community.from_census(county_fips="24510")
harmonized_nlcd_weighted = balt.harmonize(
2000,
extensive_variables=["n_total_housing_units"],
intensive_variables=["p_vacant_housing_units"],
weights_method="dasymetric",
raster=local_raster)
assert harmonized_nlcd_weighted.gdf.n_total_housing_units.sum() == 900620.0
assert_allclose(harmonized_nlcd_weighted.gdf.p_vacant_housing_units.sum(),
8832.8796,
rtol=1e-03)
def test_Community_from_census():
assert Community.from_census(state_fips="24").gdf.shape == (3759, 195)
def test_Community_from_boundary():
msas = datasets.msas()
reno = msas[msas["geoid"] == "39900"]
rn = Community.from_census(boundary=reno)
assert rn.gdf.shape == (234, 195)
def test_Community_from_stcofips():
mn = Community.from_census(state_fips="27", county_fips=["26001", "26002"])
assert mn.gdf.shape == (3881, 195)
def test_Community_from_cbsa():
la = Community.from_census(msa_fips="31080")
assert la.gdf.shape == (7683, 195)
from geosnap import Community, io
import numpy as np
reno = Community.from_census(msa_fips="39900")
columns = ["median_household_income", "p_poverty_rate", "p_unemployment_rate"]
# Aspatial Clusters
def test_gm():
r = reno.cluster(columns=columns,
method="gaussian_mixture",
best_model=True)
assert len(r.gdf.gaussian_mixture.unique()) >= 5
def test_ward():
r = reno.cluster(columns=columns, method="ward")
assert len(r.gdf.ward.unique()) == 7
def test_spectral():
r = reno.cluster(columns=columns, method="spectral")
assert len(r.gdf.spectral.unique()) == 7
def test_kmeans():
r = reno.cluster(columns=columns, method="kmeans")
from geosnap import Community
import numpy as np
from geosnap import DataStore
from numpy.testing import assert_array_equal, assert_array_almost_equal
reno = Community.from_census(msa_fips="39900", datastore=DataStore())
columns = [
"median_household_income",
"p_poverty_rate",
"p_unemployment_rate",
]
# Aspatial Clusters
def test_gm():
r = reno.cluster(columns=columns, method="gaussian_mixture", best_model=True)
assert len(r.gdf.gaussian_mixture.unique()) >= 5
def test_ward():
r = reno.cluster(columns=columns, method="ward")
assert len(r.gdf.ward.unique()) == 7
def test_spectral():
r = reno.cluster(columns=columns, method="spectral")
assert len(r.gdf.spectral.unique()) == 7
#%% from geosnap import Community #%% fip = '48' out1 = '/tmp/TX_2000.wkt' out2 = '/tmp/TX_2010.wkt' #%% data = Community.from_census(state_fips='48') dataset = data.gdf[['geometry', 'geoid', 'year']] #%% D_2000 = dataset[(dataset.year == 2000)] D_2000.to_csv(out1, sep='\t', index=False, header=False) #%% D_2010 = dataset[(dataset.year == 2010)] D_2010.to_csv(out2, sep='\t', index=False, header=False)
