from sklearn.decomposition.dict_learning import DictionaryLearning as SKLModel
import lale.helpers
import lale.operators
from numpy import nan, inf
[docs]class DictionaryLearningImpl():
def __init__(self, n_components=None, alpha=1, max_iter=1000, tol=1e-08, fit_algorithm='lars', transform_algorithm='omp', transform_n_nonzero_coefs=None, transform_alpha=None, n_jobs=None, code_init=None, dict_init=None, verbose=False, split_sign=False, random_state=None, positive_code=False, positive_dict=False):
self._hyperparams = {
'n_components': n_components,
'alpha': alpha,
'max_iter': max_iter,
'tol': tol,
'fit_algorithm': fit_algorithm,
'transform_algorithm': transform_algorithm,
'transform_n_nonzero_coefs': transform_n_nonzero_coefs,
'transform_alpha': transform_alpha,
'n_jobs': n_jobs,
'code_init': code_init,
'dict_init': dict_init,
'verbose': verbose,
'split_sign': split_sign,
'random_state': random_state,
'positive_code': positive_code,
'positive_dict': positive_dict}
[docs] def fit(self, X, y=None):
self._sklearn_model = SKLModel(**self._hyperparams)
if (y is not None):
self._sklearn_model.fit(X, y)
else:
self._sklearn_model.fit(X)
return self
_hyperparams_schema = {
'$schema': 'http://json-schema.org/draft-04/schema#',
'description': 'inherited docstring for DictionaryLearning Dictionary learning',
'allOf': [{
'type': 'object',
'required': ['n_components', 'alpha', 'max_iter', 'tol', 'fit_algorithm', 'transform_algorithm', 'transform_n_nonzero_coefs', 'transform_alpha', 'n_jobs', 'code_init', 'dict_init', 'verbose', 'split_sign', 'random_state', 'positive_code', 'positive_dict'],
'relevantToOptimizer': ['n_components', 'alpha', 'max_iter', 'tol', 'fit_algorithm', 'transform_algorithm'],
'additionalProperties': False,
'properties': {
'n_components': {
'anyOf': [{
'type': 'integer',
'minimumForOptimizer': 2,
'maximumForOptimizer': 256,
'distribution': 'uniform'}, {
'enum': [None]}],
'default': None,
'description': 'number of dictionary elements to extract'},
'alpha': {
'type': 'number',
'minimumForOptimizer': 1e-10,
'maximumForOptimizer': 1.0,
'distribution': 'loguniform',
'default': 1,
'description': 'sparsity controlling parameter'},
'max_iter': {
'type': 'integer',
'minimumForOptimizer': 10,
'maximumForOptimizer': 1000,
'distribution': 'uniform',
'default': 1000,
'description': 'maximum number of iterations to perform'},
'tol': {
'type': 'number',
'minimumForOptimizer': 1e-08,
'maximumForOptimizer': 0.01,
'distribution': 'loguniform',
'default': 1e-08,
'description': 'tolerance for numerical error'},
'fit_algorithm': {
'enum': ['lars', 'cd'],
'default': 'lars',
'description': 'lars: uses the least angle regression method to solve the lasso problem'},
'transform_algorithm': {
'enum': ['lasso_lars', 'lasso_cd', 'lars', 'omp', 'threshold'],
'default': 'omp',
'description': 'Algorithm used to transform the data'},
'transform_n_nonzero_coefs': {
'XXX TODO XXX': 'int, ``0.1 * n_features`` by default',
'description': 'Number of nonzero coefficients to target in each column of the',
'enum': [None],
'default': None},
'transform_alpha': {
'anyOf': [{
'type': 'number'}, {
'enum': [None]}],
'default': None,
'description': "If `algorithm='lasso_lars'` or `algorithm='lasso_cd'`, `alpha` is the"},
'n_jobs': {
'anyOf': [{
'type': 'integer'}, {
'enum': [None]}],
'default': None,
'description': 'Number of parallel jobs to run.'},
'code_init': {
'anyOf': [{
'type': 'array',
'items': {
'type': 'array',
'items': {
'type': 'number'},
}}, {
'enum': [None]}],
'default': None,
'description': 'initial value for the code, for warm restart'},
'dict_init': {
'anyOf': [{
'type': 'array',
'items': {
'type': 'array',
'items': {
'type': 'number'},
}}, {
'enum': [None]}],
'default': None,
'description': 'initial values for the dictionary, for warm restart'},
'verbose': {
'type': 'boolean',
'default': False,
'description': 'To control the verbosity of the procedure.'},
'split_sign': {
'type': 'boolean',
'default': False,
'description': 'Whether to split the sparse feature vector into the concatenation of'},
'random_state': {
'anyOf': [{
'type': 'integer'}, {
'type': 'object'}, {
'enum': [None]}],
'default': None,
'description': 'If int, random_state is the seed used by the random number generator;'},
'positive_code': {
'type': 'boolean',
'default': False,
'description': 'Whether to enforce positivity when finding the code.'},
'positive_dict': {
'type': 'boolean',
'default': False,
'description': 'Whether to enforce positivity when finding the dictionary'},
}}, {
'XXX TODO XXX': "Parameter: transform_n_nonzero_coefs > only used by algorithm='lars' and algorithm='omp' and is overridden by alpha in the omp case"}],
}
_input_fit_schema = {
'$schema': 'http://json-schema.org/draft-04/schema#',
'description': 'Fit the model from data in X.',
'type': 'object',
'properties': {
'X': {
'type': 'array',
'items': {
'type': 'array',
'items': {
'type': 'number'},
},
'description': 'Training vector, where n_samples in the number of samples'},
'y': {
}},
}
_input_transform_schema = {
'$schema': 'http://json-schema.org/draft-04/schema#',
'description': 'Encode the data as a sparse combination of the dictionary atoms.',
'type': 'object',
'properties': {
'X': {
'type': 'array',
'items': {
'type': 'array',
'items': {
'type': 'number'},
},
'description': 'Test data to be transformed, must have the same number of'},
},
}
_output_transform_schema = {
'$schema': 'http://json-schema.org/draft-04/schema#',
'description': 'Transformed data',
'type': 'array',
'items': {
'type': 'array',
'items': {
'type': 'number'},
},
}
_combined_schemas = {
'$schema': 'http://json-schema.org/draft-04/schema#',
'description': 'Combined schema for expected data and hyperparameters.',
'type': 'object',
'tags': {
'pre': [],
'op': ['transformer'],
'post': []},
'properties': {
'hyperparams': _hyperparams_schema,
'input_fit': _input_fit_schema,
'input_transform': _input_transform_schema,
'output_transform': _output_transform_schema},
}
if (__name__ == '__main__'):
lale.helpers.validate_is_schema(_combined_schemas)
DictionaryLearning = lale.operators.make_operator(DictionaryLearningImpl, _combined_schemas)