def _balltree(*, train, test, x_predict=None, metrics, X, leaf_size=40, metric='minkowski', **kwargs):
"""
For more info visit :
https://scikit-learn.org/stable/modules/generated/sklearn.neighbors.BallTree.html#sklearn.neighbors.BallTree
"""
model = BallTree(X, leaf_size=leaf_size, metric=metric, **kwargs)
model.fit(train[0], train[1])
model_name = 'Ball Tree'
y_hat = model.predict(test[0])
if metrics == 'accuracy':
accuracy = accuracy_score(test[1], y_hat)
if metrics == 'f1':
accuracy = f1_score(test[1], y_hat)
if metrics == 'jaccard':
accuracy = jaccard_score(test[1], y_hat)
if x_predict is None:
return (model_name, accuracy, None)
y_predict = model.predict(x_predict)
return (model_name, accuracy, y_predict)
def kde_logarithmic_knn(self, X: np.ndarray, **kwargs) -> np.ndarray:
"""
Parameters
----------
X
kwargs
Returns
-------
"""
# -- constants --
_allowed_algorithms = {'BallTree', 'NearestNeighbors'}
# -- defaults --
_algorithm = kwargs.get('knn_algorithm', 'BallTree')
assert _algorithm in _allowed_algorithms, f'The algorithm name you provided: {_algorithm} is not among ' \
f'allowed knn algorithms: {_allowed_algorithms}'
_log_regularizer = kwargs.get('log_regularizer', 0.0)
_leaf_size = kwargs.get('leaf_size', 6)
_n_neighbors = kwargs.get('knn_n_neighbors', 6)
# -- code --
if _algorithm == 'BallTree':
# create the local graph
_graph = BallTree(X, leaf_size=_leaf_size)
elif _algorithm == 'NearestNeighbors':
# create the local graph
_graph = NearestNeighbors(n_neighbors=_n_neighbors, leaf_size=_leaf_size,
algorithm='ball_tree', p=2, n_jobs=os.cpu_count() - 1)
# fit graph
_graph.fit(X)
# define the query function, so to be consistent with other functions
_graph.query = lambda X, k: _graph.kneighbors(X=X, n_neighbors=k)
else:
raise ValueError('Something horrible has happened, contact the project owner!')
# -- functions --
def _log_density_estimator(in_vector: np.ndarray, k: int = _n_neighbors) -> float:
return np.log(k) - np.log(X.shape[0]) - \
X.shape[1] * np.log(_log_regularizer + _graph.query(in_vector, k=k)[0][0][-1])
# -- code --
# it might happen that the `_log_density_estimator` is negative, so you must shift it
_densities = np.array(list(map(lambda ind: _log_density_estimator(X[ind:ind+1]), range(X.shape[0]))))
return _densities + np.abs(np.min((0.0, np.min(_densities))))
def predict(self, X):
ball_tree = BallTree()
ball_tree.fit(self.cluster_centers_)
_, indexes = ball_tree.query(X)
result = []
for idx, in indexes:
result.append(self.labels_[idx])
return result
