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_Community_from_gdfs_crs():
t90 = datasets.tracts_1990()
t00 = datasets.tracts_2000()
t90 = t90.to_crs(4236)
t00 = t00.to_crs(3857)
try:
Community.from_geodataframes([t90, t00])
except AssertionError:
print("From_gdfs constructor successfully detects inconsistent crs.")
pass
def Adaptive_Choropleth_Mapper_viz(param):
write_LOG(param)
# convert year, variable to years, variables in the param
if ('years' not in param and 'year' in param):
param['years'] = [param['year']]
if ('variables' not in param and 'variable' in param):
param['variables'] = [param['variable']]
#print(param)
# select community by state_fips, msa_fips, county_fips
if ('msa_fips' in param and param['msa_fips']):
community = Community.from_ltdb(years=param['years'],
msa_fips=param['msa_fips'])
#community = Community.from_ltdb(msa_fips=param['msa_fips'])
elif ('county_fips' in param and param['county_fips']):
community = Community.from_ltdb(years=param['years'],
county_fips=param['county_fips'])
elif ('state_fips' in param and param['state_fips']):
community = Community.from_ltdb(years=param['years'],
state_fips=param['state_fips'])
#print(community.gdf)
codebook = pd.read_csv('template/conversion_table_codebook.csv')
codebook.set_index(keys='variable', inplace=True)
labels = copy.deepcopy(param['variables'])
label = 'short_name' # default
if (param['label'] == 'variable'): label = 'variable'
if (param['label'] == 'full_name'): label = 'full_name'
if (param['label'] == 'short_name'): label = 'short_name'
if (label != 'variable'):
for idx, variable in enumerate(param['variables']):
try:
codeRec = codebook.loc[variable]
labels[idx] = codeRec[label]
except:
print("variable not found in codebook. variable:", variable)
param['labels'] = labels
write_INDEX_html(param)
write_GEO_CONFIG_js(param)
write_GEO_VARIABLES_js(community, param)
write_GEO_JSON_js(community, param)
local_dir = os.path.dirname(os.path.realpath(__file__))
fname = urllib.parse.quote('index.html')
template_dir = os.path.join(local_dir, 'ACM_' + param['filename_suffix'])
url = 'file:' + os.path.join(template_dir, fname)
webbrowser.open(url)
print('Please run ' + '"ACM_' + param['filename_suffix'] + '/index.html"' +
' to your web browser.')
print('Advanced options are available in ' + '"ACM_' +
param['filename_suffix'] + '/data/GEO_CONFIG.js"')
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_Community_from_gdfs():
t90 = datasets.tracts_1990()
t90 = t90[t90.geoid.str.startswith("11")]
t00 = datasets.tracts_2000()
t00 = t00[t00.geoid.str.startswith("11")]
assert Community.from_geodataframes([t90, t00]).gdf.shape == (380, 192)
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_transition():
"""
Testing transition modeling.
"""
columbus = Community.from_ltdb(msa_fips="18140")
columns = [
"median_household_income",
"p_poverty_rate",
"p_edu_college_greater",
"p_unemployment_rate",
]
columbus1 = columbus.cluster(
columns=[
"median_household_income",
"p_poverty_rate",
"p_edu_college_greater",
"p_unemployment_rate",
],
method="ward",
)
# 1. Markov modeling
m = transition(columbus1.gdf, cluster_col="ward")
mp = np.array([
[
0.79189189, 0.00540541, 0.0027027, 0.13243243, 0.06216216,
0.00540541
],
[0.0203252, 0.75609756, 0.10569106, 0.11382114, 0.0, 0.00406504],
[0.00917431, 0.20183486, 0.75229358, 0.01834862, 0.0, 0.01834862],
[0.1959799, 0.18341709, 0.00251256, 0.61809045, 0.0, 0.0],
[0.32307692, 0.0, 0.0, 0.0, 0.66153846, 0.01538462],
[0.09375, 0.0625, 0.0, 0.0, 0.0, 0.84375],
])
np.testing.assert_allclose(m.p, mp, RTOL)
# 2. Spatial Markov modeling
np.random.seed(5)
sm = transition(columbus1.gdf, cluster_col="ward", w_type="queen")
smp = np.array([
[0.82413793, 0.0, 0.0, 0.10689655, 0.06896552, 0.0],
[0.25, 0.5, 0.125, 0.125, 0.0, 0.0],
[0.5, 0.0, 0.5, 0.0, 0.0, 0.0],
[0.23809524, 0.0952381, 0.0, 0.66666667, 0.0, 0.0],
[0.21621622, 0.0, 0.0, 0.0, 0.75675676, 0.02702703],
[0.16666667, 0.0, 0.0, 0.0, 0.0, 0.83333333],
])
np.testing.assert_allclose(sm.P[0], smp, RTOL)
def test_sequence():
"""
Testing sequence modeling.
"""
columbus = Community.from_ltdb(msa_fips="18140")
columns = [
"median_household_income",
"p_poverty_rate",
"p_edu_college_greater",
"p_unemployment_rate",
]
columbus1 = columbus.cluster(
columns=[
"median_household_income",
"p_poverty_rate",
"p_edu_college_greater",
"p_unemployment_rate",
],
method="ward",
)
# 1. Transition-orientied optimal matching
output = sequence(columbus1.gdf,
seq_clusters=5,
dist_type="tran",
cluster_col="ward")
values = np.array([3, 3, 0, 2, 3, 1])
np.testing.assert_allclose(output[1].values[0], values, RTOL)
# 2. Hamming distance
output = sequence(columbus1.gdf,
seq_clusters=5,
dist_type="hamming",
cluster_col="ward")
values = np.array([3, 3, 0, 2, 3, 2])
np.testing.assert_allclose(output[1].values[0], values, RTOL)
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_indices():
chi = Community.from_ncdb(fips=["17031", "17019"])
assert chi.gdf.shape == (6797, 78)
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
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 numpy as np
from geosnap import Community
from geosnap.analyze import sequence, transition
RTOL = 0.00001
columbus = Community.from_ltdb(msa_fips="18140")
columns = [
"median_household_income",
"p_poverty_rate",
"p_edu_college_greater",
"p_unemployment_rate",
]
columbus1 = columbus.cluster(
columns=[
"median_household_income",
"p_poverty_rate",
"p_edu_college_greater",
"p_unemployment_rate",
],
method="ward",
)
def test_transition():
"""
Testing transition modeling.
"""
# 1. Markov modeling
def Adaptive_Choropleth_Mapper_viz(param):
# convert year, variable to years, variables in the param
if ('years' not in param and 'year' in param):
param['years'] = [param['year']]
if ('variables' not in param and 'variable' in param):
param['variables'] = [param['variable']]
#print(param)
# select community by state_fips, msa_fips, county_fips
community = None
if ('msa_fips' in param and param['msa_fips']):
community = Community.from_ltdb(years=param['years'],
msa_fips=param['msa_fips'])
#community = Community.from_ltdb(msa_fips=param['msa_fips'])
elif ('county_fips' in param and param['county_fips']):
community = Community.from_ltdb(years=param['years'],
county_fips=param['county_fips'])
elif ('state_fips' in param and param['state_fips']):
community = Community.from_ltdb(years=param['years'],
state_fips=param['state_fips'])
#print(community.gdf)
# if the user enters CSV and shapefile, use the files from the user
#### This is executed when the user enter attributes in csv file and geometroy in shapefile ######################
if (community is None and 'inputCSV' in param):
community = Community()
#community.gdf = pd.read_csv(param['inputCSV'], dtype={'geoid':str})
community.gdf = param["inputCSV"]
#print(community.gdf)
geoid = community.gdf.columns[0]
#community.gdf = community.gdf.astype(str)
#print("inputCSV: " + community.gdf.geoid)
community.gdf[community.gdf.columns[0]] = community.gdf[geoid].astype(
str)
#print("community.gdf.columns[0]:", community.gdf.columns[0])
# read shape file to df_shape
#df_shape = gpd.read_file(param['shapefile'])
df_shape = param['shapefile']
df_shape = df_shape.astype(str)
#print("shapefile: " + df_shape.GEOID10)
geokey = df_shape.columns[0]
#print(geokey)
df_shape = df_shape.set_index(geokey)
# insert geometry to community.gdf
geometry = []
for index, row in community.gdf.iterrows():
tractid = row[geoid]
try:
tract = df_shape.loc[tractid]
geometry.append(shapely.wkt.loads(tract.geometry))
except KeyError:
#print("Tract ID [{}] is not found in the shape file {}".format(tractid, param['shapefile']))
geometry.append(None)
# print( "geometry" in community.gdf )
#f hasattr(community.gdf, "geoemtry"):
#if (community.gdf["geoemtry"] is None):
# pass
#else:
if (("geometry" in community.gdf) == False):
community.gdf.insert(len(community.gdf.columns), "geometry",
geometry)
################################################################################################################
community.gdf = community.gdf.replace([np.inf, -np.inf], np.nan)
# check if geometry is not null for Spatial Clustering
community.gdf = community.gdf[pd.notnull(community.gdf['geometry'])]
#print(community.gdf)
codebook = pd.read_csv('template/conversion_table_codebook.csv')
codebook.set_index(keys='variable', inplace=True)
labels = copy.deepcopy(param['variables'])
label = 'short_name' # default
if (param['label'] == 'variable'): label = 'variable'
if (param['label'] == 'full_name'): label = 'full_name'
if (param['label'] == 'short_name'): label = 'short_name'
if (label != 'variable'):
for idx, variable in enumerate(param['variables']):
try:
codeRec = codebook.loc[variable]
labels[idx] = codeRec[label]
except:
print("variable not found in codebook. variable:", variable)
param['labels'] = labels
write_INDEX_html(param)
write_CONFIG_js(param)
write_VARIABLES_js(community, param)
write_GEO_JSON_js(community, param)
'''
#Create directory for local machine
local_dir = os.path.dirname(os.path.realpath(__file__))
fname =urllib.parse.quote('index.html')
template_dir = os.path.join(local_dir, 'ACM_' + param['filename_suffix'])
url = 'file:' + os.path.join(template_dir, fname)
webbrowser.open(url)
print('Please run ' + '"ACM_' + param['filename_suffix']+'/index.html"'+' to your web browser.')
print('Advanced options are available in ' + '"ACM_' + param['filename_suffix']+'/data/CONFIG.js"')
'''
#Create directory for Visualization
servers = list(notebookapp.list_running_servers())
servers1 = 'https://cybergisx.cigi.illinois.edu' + servers[0][
"base_url"] + 'view'
servers2 = 'https://cybergisx.cigi.illinois.edu' + servers[0][
"base_url"] + 'edit'
cwd = os.getcwd()
prefix_cwd = "/home/jovyan/work"
cwd = cwd.replace(prefix_cwd, "")
# This is for Jupyter notebbok installed in your PC
local_dir1 = cwd
local_dir2 = cwd
#This is for CyberGISX. Uncomment two command lines below when you run in CyberGIX Environment
#local_dir1 = servers1 + cwd
#local_dir2 = servers2 + cwd
#print(local_dir)
fname = urllib.parse.quote('index.html')
template_dir = os.path.join(local_dir1, 'ACM_' + param['filename_suffix'])
#url = 'file:' + os.path.join(template_dir, fname)
url = os.path.join(template_dir, fname)
webbrowser.open(url)
print(
'To see your visualization, click the URL below (or locate the files):'
)
print(url)
print('Advanced options are available in ')
print(local_dir2 + '/' + 'ACM_' + param['filename_suffix'] +
'/data/CONFIG_' + param['filename_suffix'] + '.js')
def test_Community_from_lodes():
dc = Community.from_lodes(state_fips="11", years=[2010, 2015])
assert dc.gdf.shape == (6046, 57)
def test_Community_from_lodes():
dc = Community.from_lodes(state_fips="10", years=[2008, 2015])
assert dc.gdf.shape == (8674, 57)
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")
def test_Community_from_lodes():
de = Community.from_lodes(state_fips="10", years=[2008, 2015])
assert de.gdf.shape == (41598, 58)
#%% 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)