def test_plot_spatial_weights():
# get data
gdf = gpd.read_file(examples.get_path('43MUE250GC_SIR.shp'))
gdf.head()
# calculate weights
weights = Queen.from_dataframe(gdf)
# plot weights
fig, _ = plot_spatial_weights(weights, gdf)
plt.close(fig)
# calculate nonplanar_joins
wnp = libpysal.weights.util.nonplanar_neighbors(weights, gdf)
# plot new joins
fig2, _ = plot_spatial_weights(wnp, gdf)
plt.close(fig2)
#customize
fig3, _ = plot_spatial_weights(wnp,
gdf,
nonplanar_edge_kws=dict(color='#4393c3'))
plt.close(fig3)
def test_plot_spatial_weights():
# get data
gdf = gpd.read_file(examples.get_path('43MUE250GC_SIR.shp'))
gdf.head()
# calculate weights
weights = Queen.from_dataframe(gdf)
# plot weights
fig, _ = plot_spatial_weights(weights, gdf)
plt.close(fig)
# calculate nonplanar_joins
wnp = libpysal.weights.util.nonplanar_neighbors(weights, gdf)
# plot new joins
fig2, _ = plot_spatial_weights(wnp, gdf)
plt.close(fig2)
#customize
fig3, _ = plot_spatial_weights(wnp,
gdf,
nonplanar_edge_kws=dict(color='#4393c3'))
plt.close(fig3)
# plot in existing figure
fig4, axs = plt.subplots(1, 3)
plot_spatial_weights(wnp, gdf, ax=axs[0])
plt.close(fig4)
# uses a column as the index for spatial weights object
weights_index = Queen.from_dataframe(gdf, idVariable="CD_GEOCMU")
fig, _ = plot_spatial_weights(weights_index, gdf, indexed_on="CD_GEOCMU")
plt.close(fig)
from libpysal.weights.contiguity import Queen
import libpysal
from libpysal import examples
import matplotlib.pyplot as plt
import geopandas as gpd
from splot.libpysal import plot_spatial_weights
examples.explain('tokyo')
gdf = gpd.read_file(examples.get_path('tokyomet262.shp'))
gdf.head()
weight = Queen.from_dataframe(gdf)
plot_spatial_weights(weight, gdf)
plt.show()
wnp = libpysal.weights.util.nonplanar_neighbors(weight, gdf)
plot_spatial_weights(wnp, gdf)
plt.show()
plot_spatial_weights(wnp, gdf, nonplanar_edge_kws=dict(color='#4393c3'))
plt.show()
def multiscalar(geoDF,populationCensusTif):
geoDF=geoDF.to_crs(communityArea_CRS)
plt.figure()
geoDF.plot(column='Gonorrhea in Females',cmap='OrRd',figsize=(10, 10),legend=True,scheme='quantiles') # scheme='quantiles'
# geoplot.polyplot(geoDF,figsize=(20, 20))
# pc=gpd.read_file(populationCensus)
# print(pc.columns)
# print(pc.crs)
# pc.to_file(os.path.join(dataRoot,"populationCensus_save.shp"))
# Feature_to_Raster(os.path.join(dataRoot,"populationCensus_save.shp"), os.path.join(dataRoot,"popuC_1.tif"), 20, field_name=['CENSUS_BLO'], NoData_value=-9999)
geoDF_sample=zonal_stats(geoDF,populationCensusTif,stats="count min mean max median")
geoDF_sample_df=pd.DataFrame(geoDF_sample)
geoDF_sample_df.rename(columns={"count":"popuCount", "min":"popuMin", "mean":"popuMean", "max":"popuMax", "median":"popuMedian"},inplace=True)
geoDF_merge=geoDF.merge(geoDF_sample_df,left_on=geoDF.index, right_on=geoDF_sample_df.index)
groups_list = ['Cancer (All Sites)','Lung Cancer','Tuberculosis','Gonorrhea in Females']
geoDF_merge[groups_list]=geoDF_merge[groups_list].fillna(0)
geoDF_merge["sumGroup"]=geoDF_merge[groups_list].sum(axis=1)
for i in range(len(groups_list)):
geoDF_merge['comp_' + groups_list[i]] = geoDF_merge[groups_list[i]] /geoDF_merge["sumGroup"] #input_df['popuMean']
df_pts=geoDF_merge.copy()
df=df_pts
w_queen = Queen.from_dataframe(df)
w_rook = Rook.from_dataframe(df)
w_kernel_1k = Kernel.from_dataframe(df_pts, bandwidth=2500)
w_kernel_2k = Kernel.from_dataframe(df_pts, bandwidth=3000)
#01-show spatial weights structure
fig, ax = plt.subplots(1,4, figsize=(16,4))
plot_spatial_weights(w_queen, df, ax=ax[0])
ax[0].set_title('queen')
plot_spatial_weights(w_rook, df, ax=ax[1])
ax[1].set_title('rook')
plot_spatial_weights(w_kernel_1k, df, ax=ax[2])
ax[2].set_title('kernel 1k')
plot_spatial_weights(w_kernel_2k, df, ax=ax[3])
ax[3].set_title('kernel 2k')
#02-show local environment
def plot_local_environment(w, ax):
from segregation.spatial.spatial_indexes import _build_local_environment
d = _build_local_environment(df, groups_list, w)
d['geometry'] = df.geometry
d = gpd.GeoDataFrame(d)
d.plot('Lung Cancer', k=6, scheme='quantiles', ax=ax)
ax.axis('off')
plt.figure()
fig, axs = plt.subplots(1,4, figsize=(16,4))
for i, wtype in enumerate([w_queen, w_rook, w_kernel_1k, w_kernel_2k]):
plot_local_environment(w=wtype, ax=axs[i])
#03-different local environments result in different segregation statistics
#aspatial
multiInfo=MultiInformationTheory(df, groups_list).statistic
print("aspatial:",multiInfo)
#rook neighborhood
rookNeighborhood=SpatialInformationTheory(df, groups_list, w=w_rook).statistic
print("rook neighborhood:",rookNeighborhood)
#queen neighborhood
queenNeighbor=SpatialInformationTheory(df, groups_list, w=w_queen).statistic
print("queen neighborhood:",queenNeighbor)
# 1 kilometer kernel distance neighborhood
kernelDisNeighbor_a=SpatialInformationTheory(df, groups_list, w=w_kernel_1k).statistic
print( "kernel distance neighborhood_A:",kernelDisNeighbor_a)
# 2 kilometer kernel distance neighborhood
kernelDisNeighbor_b=SpatialInformationTheory(df, groups_list, w=w_kernel_2k).statistic
print("kernel distance neighborhood_B:",kernelDisNeighbor_b)
distances = [1000.,2000.,3000.,4000.,5000.] # note these are floats [1000.,2000.,3000.,4000.,5000.]
euclidian_profile = compute_segregation_profile(df_pts, groups=groups_list, distances=distances)
print("euclidian_profile",euclidian_profile)
#04-local street network can have a big impact on -
df = df.to_crs(epsg=4326)
# net = get_osm_network(df)
# net.save_hdf5('dc_network.h5') #it can take awhile to download a street network, so you can save and read it back in using pandana
dc_networkPath=r"C:\Users\richi\omen-richiebao\omen-code\Chicago_code\Chicago Health_spatioTemporalAnalysis\data\dc_network.h5"
net = Network.from_hdf5(dc_networkPath)
# net = Network.from_hdf5('dc_network.h5')
#three different segregation profiles
network_linear_profile = compute_segregation_profile(df_pts, groups=groups_list, network=net, distances=distances)
network_exponential_profile = compute_segregation_profile(df_pts, groups=groups_list, network=net, distances=distances, decay='exp', precompute=False)
plt.figure()
fig, ax = plt.subplots(figsize=(12,8))
ax.scatter(euclidian_profile.keys(), euclidian_profile.values(), c='green', label='euclidian exp')
ax.plot(list(euclidian_profile.keys()), list(euclidian_profile.values()), c='green') #TypeError: float() argument must be a string or a number, not 'dict_keys'
ax.scatter(network_linear_profile.keys(), network_linear_profile.values(), c='red', label='net linear')
ax.plot(list(network_linear_profile.keys()), list(network_linear_profile.values()), c='red')
ax.scatter(network_exponential_profile.keys(), network_exponential_profile.values(), c='blue', label='net exp')
ax.plot(list(network_exponential_profile.keys()), list(network_exponential_profile.values()), c='blue')
plt.xlabel('meters')
plt.ylabel('SIT')
plt.legend()
plt.show()
tess_c = gpd.read_file(path, layer="tess_c")
blg = pd.concat([blg_s, blg_c])
tess = pd.concat([tess_s, tess_c])
blg = blg.sort_values("uID")
blg.reset_index(inplace=True)
tess = tess.loc[tess["uID"].isin(blg["uID"])]
tess = tess.sort_values("uID")
tess.reset_index(inplace=True)
weights = libpysal.weights.contiguity.Queen.from_dataframe(tess)
f, ax = plt.subplots(figsize=(20, 10))
tess.plot(ax=ax)
plot_spatial_weights(weights, blg, ax=ax)
#plt.savefig(
# "contiguity_diagram.svg",
# dpi=300,
# bbox_inches="tight",
#)
# In[ ]: