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)
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
import os
import pytest
from geosnap import DataStore, io
try:
LTDB = os.environ["LTDB_SAMPLE"]
NCDB = os.environ["NCDB"]
except:
LTDB = None
NCDB = None
store = DataStore()
def test_nces_schools():
schools = io.get_nces(store, dataset="schools")
assert schools.shape == (102209, 26)
def test_nces_school_dists():
dists = io.get_nces(store, dataset="school_districts")
assert dists.shape == (13352, 18)
def test_ejscreen():
ej = io.get_ejscreen(store, years=[2018], fips=["11"])
assert ej.shape == (450, 369)
def test_nces_sabs():
def test_linc_method():
columns = [
"median_household_income",
"p_poverty_rate",
"p_unemployment_rate",
"n_total_pop",
]
reno = get_census(DataStore(), msa_fips="39900")
rdf = harmonize(reno, target_year=2010, intensive_variables=columns)
_, model = analyze.cluster(rdf,
columns=columns,
method="ward",
return_model=True)
l = model.lincs.linc.values
assert_array_almost_equal(
l,
np.array([
0.9047619,
0.94594595,
0.82608696,
0.875,
0.97142857,
0.9047619,
1.0,
0.96428571,
0.97560976,
1.0,
0.82608696,
1.0,
0.92682927,
0.94285714,
1.0,
0.94285714,
0.92682927,
1.0,
0.90909091,
0.94285714,
1.0,
1.0,
1.0,
0.975,
0.9047619,
0.97560976,
1.0,
0.82608696,
0.82608696,
0.94594595,
0.875,
0.875,
0.96428571,
0.875,
0.90625,
1.0,
0.9137931,
0.98360656,
1.0,
0.875,
1.0,
0.98181818,
0.97619048,
0.90909091,
0.98181818,
0.90909091,
0.94594595,
0.82608696,
0.97619048,
0.90909091,
0.90625,
0.9137931,
0.93333333,
0.93333333,
1.0,
1.0,
0.93333333,
0.93333333,
0.975,
0.90625,
0.96666667,
0.96666667,
0.98507463,
0.9137931,
0.94339623,
0.93939394,
0.93939394,
0.94339623,
0.94339623,
0.9137931,
0.97142857,
0.875,
0.93939394,
0.93939394,
0.93939394,
0.98507463,
1.0,
1.0,
0.9047619,
0.96666667,
0.9047619,
0.90909091,
0.94339623,
0.90625,
0.90625,
0.9137931,
0.9137931,
0.98214286,
0.984375,
0.95918367,
0.95918367,
0.95918367,
0.92682927,
0.92682927,
0.98360656,
0.96551724,
0.98214286,
0.96551724,
0.984375,
1.0,
1.0,
0.98214286,
0.96551724,
0.90625,
0.90625,
0.98214286,
1.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
]),
decimal=3,
)
def test_linc_from_gdf():
columns = [
"median_household_income",
"p_poverty_rate",
"p_unemployment_rate",
]
reno = get_census(DataStore(), msa_fips="39900")
rdf = harmonize(reno, target_year=1990, intensive_variables=columns)
rdf = analyze.cluster(reno, columns=columns, method="ward")
l = lincs_from_gdf(rdf,
unit_index="geoid",
temporal_index="year",
cluster_col="ward")
assert_array_almost_equal(
l.linc.values,
np.array([
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.85714286,
0.5,
1.0,
0.8,
0.0,
0.0,
0.0,
1.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.5,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
1.0,
0.8,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.85714286,
0.5,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
1.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.5,
0.0,
1.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.5,
1.0,
1.0,
1.0,
0.0,
0.5,
0.5,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
0.0,
]),
decimal=3,
)
def test_linc_from_gdf_subset():
columns = [
"median_household_income",
"p_poverty_rate",
"p_unemployment_rate",
"n_total_pop",
]
reno = get_census(DataStore(), msa_fips="39900")
rdf = harmonize(reno, target_year=1990, intensive_variables=columns)
rdf = analyze.cluster(
rdf,
columns=columns,
method="ward",
)
l = lincs_from_gdf(
rdf,
unit_index="geoid",
temporal_index="year",
cluster_col="ward",
periods=[2000, 2010],
)
assert_array_almost_equal(
l.linc.values,
np.array([
0.96969697,
0.78571429,
0.8,
0.75,
0.66666667,
0.8125,
0.78571429,
0.80952381,
1.0,
0.8,
0.75,
0.74074074,
0.80952381,
0.80952381,
0.92307692,
1.0,
0.8,
0.78571429,
0.78571429,
0.75,
0.8125,
0.75,
0.74074074,
0.74074074,
0.8,
0.75,
0.66666667,
0.90909091,
0.66666667,
0.92307692,
1.0,
1.0,
0.74074074,
0.80952381,
1.0,
1.0,
1.0,
0.74074074,
0.96969697,
1.0,
0.8125,
0.74074074,
0.74074074,
1.0,
0.80952381,
0.8125,
0.96153846,
0.90909091,
0.74074074,
0.66666667,
0.66666667,
0.66666667,
0.66666667,
0.66666667,
0.66666667,
0.96153846,
0.66666667,
0.66666667,
]),
decimal=3,
)
from geosnap import DataStore
datasets = DataStore()
def test_data_dir():
loc = datasets.show_data_dir()
assert len(loc) > 5
def test_acs():
df = datasets.acs(year=2012, states=["11"])
assert df.shape == (179, 104)
def test_tracts90():
df = datasets.tracts_1990(states=["11"])
assert df.shape == (192, 164)
def test_tracts00():
df = datasets.tracts_2000(states=["11"])
assert df.shape == (188, 192)
def test_tracts10():
df = datasets.tracts_2010(states=["11"])
assert df.shape == (179, 194)
def test_counties():
assert datasets.counties().shape == (3233, 2)