def predict(self, X, **kwargs):
""" Predict rankings for a given collection of sets of objects.
Parameters
----------
X : dict or numpy array
Dictionary with a mapping from ranking size to numpy arrays
or a single numpy array of size:
(n_instances, n_objects, n_features)
Returns
-------
Y : dict or numpy array
Dictionary with a mapping from ranking size to numpy arrays
or a single numpy array of size:
(n_instances, n_objects)
Predicted ranking
"""
self.logger.debug('Predicting started')
predicted_scores = self.predict_scores(X, **kwargs)
self.logger.debug('Predicting scores complete')
predicted_rankings = scores_to_rankings(predicted_scores)
self.logger.debug('Predicting ranks complete')
del predicted_scores
return predicted_rankings
def make_similarity_based_dataset(self, n_instances, seed=42):
random_state = np.random.RandomState(seed=seed)
X = np.empty((n_instances, self.n_objects, self.n_features),
dtype=float)
rankings = np.empty((n_instances, self.n_objects), dtype=int)
similarity_scores = np.empty_like(rankings, dtype=float)
movie_features = self.movies_df.as_matrix()[:, 4:]
for i in range(n_instances):
subset = random_state.choice(movie_features.shape[0],
size=self.n_objects,
replace=False)
X[i] = movie_features[subset]
query = random_state.choice(self.n_objects, size=1)
one_row = [
self.similarity_matrix_lin_list[get_key_for_indices(i, j)]
for i, j in product(subset[query], subset)
]
similarity_scores[i] = np.array(one_row)
# Higher the similarity lower the rank of the object
rankings = scores_to_rankings(similarity_scores)
for i, x in enumerate(X):
x = StandardScaler().fit_transform(x)
X[i] = x
return X, rankings
def make_gp_transitive(self,
n_instances=1000,
n_objects=5,
noise=0.0,
n_features=100,
kernel_params=None,
seed=42,
**kwd):
"""Creates a nonlinear object ranking problem by sampling from a
Gaussian process as the latent utility function.
Note that this function needs to compute a kernel matrix of size
(n_instances * n_objects) ** 2, which could allocate a large chunk of the
memory."""
random_state = np.random.RandomState(seed=seed)
if kernel_params is None:
kernel_params = dict()
n_total = n_instances * n_objects
X = random_state.rand(n_total, n_features)
L = np.linalg.cholesky(Matern(**kernel_params)(X))
f = (L.dot(random_state.randn(n_total)) +
random_state.normal(scale=noise, size=n_total))
X = X.reshape(n_instances, n_objects, n_features)
f = f.reshape(n_instances, n_objects)
rankings = scores_to_rankings(f)
return X, rankings
def make_similarity_based_dataset(self, datatype='train', seed=42):
"""Picks a random subset of objects, determines the medoid and ranks the objects
based on the distance to the medoid.
The medoid is also included in the ordering."""
random_state = np.random.RandomState(seed=seed)
if datatype == 'train':
image_features = self.image_features_train
n_instances = self.n_train_instances
similarity_matrix_file = self.similarity_matrix_train_file
elif datatype == 'test':
image_features = self.image_features_test
n_instances = self.n_test_instances
similarity_matrix_file = self.similarity_matrix_test_file
X = np.empty((n_instances, self.n_objects, self.n_features), dtype=float)
similarity_scores = np.empty((n_instances, self.n_objects), dtype=float)
similarity_matrix_lin_list = initialize_similarity_matrix(similarity_matrix_file)
for i in range(n_instances):
subset = random_state.choice(image_features.shape[0], size=self.n_objects, replace=False)
X[i] = image_features[subset]
query = random_state.choice(self.n_objects, size=1)
one_row = [similarity_matrix_lin_list[get_key_for_indices(i, j)] for i, j in product(subset[query], subset)]
similarity_scores[i] = np.array(one_row)
rankings = scores_to_rankings(similarity_scores)
for i, x in enumerate(X):
x = StandardScaler().fit_transform(x)
X[i] = x
return X, rankings
def predict(self, X, **kwargs):
self.logger.info("Predicting ranks")
if isinstance(X, dict):
result = dict()
for n, scores in self.predict_scores(X, **kwargs).items():
predicted_rankings = scores_to_rankings(scores)
result[n] = predicted_rankings
return result
return ObjectRanker.predict(self, X, **kwargs)
def make_linear_transitive(self,
n_instances=1000,
n_objects=5,
noise=0.0,
n_features=100,
n_informative=10,
seed=42,
**kwd):
random_state = np.random.RandomState(seed=seed)
X, y, coeff = make_regression(n_samples=n_instances * n_objects,
n_features=n_features,
n_informative=n_informative,
coef=True,
noise=noise,
random_state=random_state)
X = X.reshape(n_instances, n_objects, n_features)
y = y.reshape(n_instances, n_objects)
rankings = scores_to_rankings(y)
return X, rankings
def predict(self, X, **kwargs):
""" Predict orderings for the labels for a given context vector.
Parameters
----------
X : array-like, shape (n_samples, n_features)
Returns
-------
Y : array-like, shape (n_samples, n_labels)
Predicted orderings
"""
self.logger.debug('Predicting started')
predicted_scores = self.predict_scores(X, **kwargs)
self.logger.debug('Predicting scores complete')
predicted_orderings = scores_to_rankings(predicted_scores)
self.logger.debug('Predicting ranks complete')
del predicted_scores
return predicted_orderings
def predict(self, Xo, Xc, **kwargs):
s = self.predict_scores(Xo, Xc, **kwargs)
return scores_to_rankings(s)
