def test_PointAdaptive_kNN(data_Fig1, output_Fig1_labels,
output_Fig1_labelsHalos, output_Fig1_borders):
est = DensityPeakAdvanced(Z=1.5, n_jobs=-1)
assert est.dim == None
assert est.k_max == 1000
assert est.D_thr == 23.92812698
assert est.metric == "euclidean"
assert est.dim_algo == "twoNN"
est.fit(data_Fig1)
assert hasattr(est, 'is_fitted_')
assert est.k_max_ == max(est.k_hat_)
print(len(data_Fig1), len(est.densities_))
assert len(data_Fig1) == len(est.densities_)
assert_array_equal(est.labels_, [c - 1 for c in output_Fig1_labels["clu"]])
is_almost_equal(est.halos_,
[c - 1 for c in output_Fig1_labelsHalos["clu"]], 0.0, 0)
#assert_array_equal(est.halos_, output_Fig1_labelsHalos["clu"])
assert_array_equal(
[est.topography_[i][0] + 1 for i in range(len(est.topography_))],
output_Fig1_borders["i"])
assert_array_equal(
[est.topography_[i][1] + 1 for i in range(len(est.topography_))],
output_Fig1_borders["j"])
npt.assert_almost_equal(
[est.topography_[i][2] for i in range(len(est.topography_))],
output_Fig1_borders["rho_b"],
decimal=3)
npt.assert_almost_equal(
[est.topography_[i][3] for i in range(len(est.topography_))],
output_Fig1_borders["err_rho_b"],
decimal=3)
def calculate_kl(p, files, fine_grid, n, size):
kls = []
for file in files:
x = np.load(file)
np.random.shuffle(x)
x = x[:size, :n]
print(x.shape)
nn = NearestNeighbors(n_neighbors=3).fit(x)
dens = DensityPeakAdvanced(k_max=500).fit(x)
q = predict(dens, nn, fine_grid)
kl = JS(p, q)
kls.append(kl)
print(file, kl)
return kls
def ref_prob(filename, grid, n, size, D_thr=15):
x = np.load(filename)[:size, :n]
nn = NearestNeighbors(n_neighbors=3).fit(x)
dens = DensityPeakAdvanced(D_thr=D_thr, k_max=500).fit(x)
return predict(dens, nn, grid), x
def test_metric_fail():
with pytest.raises(ValueError):
_ = DensityPeakAdvanced(metric='a_metric')
def test_metric_callable():
_ = DensityPeakAdvanced(metric=lambda x, y: 1)
import pytest
from sklearn.utils.estimator_checks import check_estimator
from sklearn.utils.estimator_checks import (
_construct_instance, _set_checking_parameters, _get_check_estimator_ids,
check_parameters_default_constructible,
check_class_weight_balanced_linear_classifier, parametrize_with_checks)
from Pipeline.DPA import PointAdaptive_kNN
from Pipeline.DPA import twoNearestNeighbors
from Pipeline.DPA import DensityPeakAdvanced
@parametrize_with_checks(
[DensityPeakAdvanced(),
PointAdaptive_kNN(),
twoNearestNeighbors()])
def test_scikitlearn_compatible_estimator(estimator, check):
check(estimator)