def test_kde_algorithm_metric_choice():
"""Smoke test for various metrics and algorithms"""
np.random.seed(0)
X = np.random.random((10, 2)) # 2 features required for haversine dist.
Y = np.random.random((10, 2))
for algorithm in ["auto", "ball_tree", "kd_tree"]:
for metric in ["euclidean", "minkowski", "manhattan", "chebyshev", "haversine"]:
if algorithm == "kd_tree" and metric not in KDTree.valid_metrics:
assert_raises(ValueError, KernelDensity, algorithm=algorithm, metric=metric)
else:
kde = KernelDensity(algorithm=algorithm, metric=metric)
kde.fit(X)
y_dens = kde.eval(Y)
assert_equal(y_dens.shape, Y.shape[:1])
def test_kde_algorithm_metric_choice():
"""Smoke test for various metrics and algorithms"""
np.random.seed(0)
X = np.random.random((10, 2)) # 2 features required for haversine dist.
Y = np.random.random((10, 2))
for algorithm in ['auto', 'ball_tree', 'kd_tree']:
for metric in [
'euclidean', 'minkowski', 'manhattan', 'chebyshev', 'haversine'
]:
if algorithm == 'kd_tree' and metric not in KDTree.valid_metrics:
assert_raises(ValueError,
KernelDensity,
algorithm=algorithm,
metric=metric)
else:
kde = KernelDensity(algorithm=algorithm, metric=metric)
kde.fit(X)
y_dens = kde.eval(Y)
assert_equal(y_dens.shape, Y.shape[:1])
subplots = (211, 212)
k_values = (10, 100)
for N, k, subplot in zip(N_values, k_values, subplots):
ax = fig.add_subplot(subplot)
xN = x[:N]
t = np.linspace(-10, 30, 1000)
# Compute density with KDE
if use_sklearn_KDE:
kde = KernelDensity(0.1, kernel='gaussian')
kde.fit(xN[:, None])
dens_kde = np.exp(kde.score_samples(t[:, None]))
else:
kde = KDE('gaussian', h=0.1).fit(xN[:, None])
dens_kde = kde.eval(t[:, None]) / N
# Compute density with Bayesian nearest neighbors
nbrs = KNeighborsDensity('bayesian', n_neighbors=k).fit(xN[:, None])
dens_nbrs = nbrs.eval(t[:, None]) / N
# plot the results
ax.plot(t,
true_pdf(t),
':',
color='black',
zorder=3,
label="Generating Distribution")
ax.plot(xN, -0.005 * np.ones(len(xN)), '|k')
hist(xN,
bins='blocks',
# Plot map of South America with distributions of each species
fig = plt.figure()
fig.subplots_adjust(left=0.05, right=0.95, wspace=0.05)
for i in range(2):
plt.subplot(1, 2, i + 1)
# construct a kernel density estimate of the distribution
print(" - computing KDE in spherical coordinates")
kde = KernelDensity(bandwidth=0.04, metric='haversine',
kernel='gaussian', algorithm='ball_tree')
kde.fit(Xtrain[ytrain == i])
# evaluate only on the land: -9999 indicates ocean
Z = -9999 + np.zeros(land_mask.shape[0])
Z[land_mask] = np.exp(kde.eval(xy))
Z = Z.reshape(X.shape)
# plot contours of the density
levels = np.linspace(0, Z.max(), 25)
plt.contourf(X, Y, Z, levels=levels, cmap=plt.cm.Reds)
if basemap:
print(" - plot coastlines using basemap")
m = Basemap(projection='cyl', llcrnrlat=Y.min(),
urcrnrlat=Y.max(), llcrnrlon=X.min(),
urcrnrlon=X.max(), resolution='c')
m.drawcoastlines()
m.drawcountries()
else:
print(" - plot coastlines from coverage")
subplots = (211, 212)
k_values = (10, 100)
for N, k, subplot in zip(N_values, k_values, subplots):
ax = fig.add_subplot(subplot)
xN = x[:N]
t = np.linspace(-10, 30, 1000)
# Compute density with KDE
if use_sklearn_KDE:
kde = KernelDensity(0.1, kernel='gaussian')
kde.fit(xN[:, None])
dens_kde = np.exp(kde.score_samples(t[:, None]))
else:
kde = KDE('gaussian', h=0.1).fit(xN[:, None])
dens_kde = kde.eval(t[:, None]) / N
# Compute density with Bayesian nearest neighbors
nbrs = KNeighborsDensity('bayesian', n_neighbors=k).fit(xN[:, None])
dens_nbrs = nbrs.eval(t[:, None]) / N
# plot the results
ax.plot(t, true_pdf(t), ':', color='black', zorder=3,
label="Generating Distribution")
ax.plot(xN, -0.005 * np.ones(len(xN)), '|k')
hist(xN, bins='blocks', ax=ax, normed=True, zorder=1,
histtype='stepfilled', color='k', alpha=0.2,
label="Bayesian Blocks")
ax.plot(t, dens_nbrs, '-', lw=1.5, color='gray', zorder=2,
label="Nearest Neighbors (k=%i)" % k)
ax.plot(t, dens_kde, '-', color='black', zorder=3,
fig.subplots_adjust(left=0.05, right=0.95, wspace=0.05)
for i in range(2):
plt.subplot(1, 2, i + 1)
# construct a kernel density estimate of the distribution
print " - computing KDE in spherical coordinates"
kde = KernelDensity(bandwidth=0.04,
metric='haversine',
kernel='gaussian',
algorithm='ball_tree')
kde.fit(Xtrain[ytrain == i])
# evaluate only on the land: -9999 indicates ocean
Z = -9999 + np.zeros(land_mask.shape[0])
Z[land_mask] = np.exp(kde.eval(xy))
Z = Z.reshape(X.shape)
# plot contours of the density
levels = np.linspace(0, Z.max(), 25)
plt.contourf(X, Y, Z, levels=levels, cmap=plt.cm.Reds)
if basemap:
print " - plot coastlines using basemap"
m = Basemap(projection='cyl',
llcrnrlat=Y.min(),
urcrnrlat=Y.max(),
llcrnrlon=X.min(),
urcrnrlon=X.max(),
resolution='c')
m.drawcoastlines()
