def init_od_mahalanobis(state_dict: Dict) -> Mahalanobis:
"""
Initialize Mahalanobis.
Parameters
----------
state_dict
Dictionary containing the parameter values.
Returns
-------
Initialized Mahalanobis instance.
"""
od = Mahalanobis(threshold=state_dict['threshold'],
n_components=state_dict['n_components'],
std_clip=state_dict['std_clip'],
start_clip=state_dict['start_clip'],
max_n=state_dict['max_n'],
cat_vars=state_dict['cat_vars'],
ohe=state_dict['ohe'])
od.d_abs = state_dict['d_abs']
od.clip = state_dict['clip']
od.mean = state_dict['mean']
od.C = state_dict['C']
od.n = state_dict['n']
return od
def test_mahalanobis(mahalanobis_params):
threshold, n_components, std_clip, start_clip, max_n, \
threshold_perc, return_instance_score = mahalanobis_params
X, y = load_iris(return_X_y=True)
mh = Mahalanobis(threshold,
n_components=n_components,
std_clip=std_clip,
start_clip=start_clip,
max_n=max_n)
assert mh.threshold == threshold
assert mh.n == 0
assert mh.meta == {
'name': 'Mahalanobis',
'detector_type': 'online',
'data_type': 'tabular',
'version': __version__
}
mh.infer_threshold(X, threshold_perc=threshold_perc)
assert mh.n == X.shape[0]
iscore = mh.score(X) # noqa
assert mh.n == 2 * X.shape[0]
assert mh.mean.shape[0] == X.shape[1]
assert mh.C.shape == (X.shape[1], X.shape[1])
assert (np.diag(mh.C) >= 0).all()
od_preds = mh.predict(X, return_instance_score=return_instance_score)
assert mh.n == 3 * X.shape[0]
assert od_preds['meta'] == mh.meta
if return_instance_score:
assert od_preds['data']['is_outlier'].sum() == (
od_preds['data']['instance_score'] >
mh.threshold).astype(int).sum()
else:
assert od_preds['data']['instance_score'] is None
])
threshold_net = tf.keras.Sequential([
InputLayer(input_shape=(seq_len, latent_dim)),
Dense(5, activation=tf.nn.relu)
])
# define model
inputs = tf.keras.Input(shape=(input_dim, ))
outputs = tf.keras.layers.Dense(2, activation=tf.nn.softmax)(inputs)
model = tf.keras.Model(inputs=inputs, outputs=outputs)
detector = [
AdversarialAE(threshold=threshold, model=model, **kwargs),
IForest(threshold=threshold),
Mahalanobis(threshold=threshold),
OutlierAEGMM(threshold=threshold,
gmm_density_net=gmm_density_net,
n_gmm=n_gmm,
**kwargs),
OutlierVAE(threshold=threshold,
latent_dim=latent_dim,
samples=samples,
**kwargs),
OutlierAE(threshold=threshold, **kwargs),
OutlierVAEGMM(threshold=threshold,
gmm_density_net=gmm_density_net,
n_gmm=n_gmm,
latent_dim=latent_dim,
samples=samples,
**kwargs),