prevision_quantum_nn.preprocessing.preprocess
¶
preprocessing module
Module Contents¶
Classes¶
Class Preprocessor, preprocesses features for quantum models. |
-
class
prevision_quantum_nn.preprocessing.preprocess.
Preprocessor
(params)¶ Class Preprocessor, preprocesses features for quantum models.
-
polynomial_expander
¶ the polynomial expander
- Type
-
polynomial_expansion_type
¶ can be either polynomial_features or kronecker
- Type
str
-
feature_engineer
¶ feature engineer object
- Type
-
dimension_reduction_fitter
¶ can be either wrapper or pca
- Type
str
-
force_dimension_reduction
¶ if True, dimension reduction will be forced if False, errors will prompt if the data does not fit into the quantum computer
- Type
bool
-
__getstate__
(self)¶
-
__setstate__
(self, d)¶
-
check_preprocessor
(self)¶ Checks the preprocessor consistency.
- Raises
ValueError if parameters are not consistent –
-
build_for_model
(self, architecture_type, encoding, num_q, type_problem, rl_bounds=None, rl_tanh_mask=None, rl_discrete_depth=None)¶ Builds the preprocessor for a given model.
- Parameters
model (QuantumNeuralNetwork) – model that needs to be used with this preprocessing
-
fit_transform
(self, features, labels)¶ Fit and transforms features.
- Parameters
features (numpy array) – input features
labels (numpy array) – data label
- Returns
- numpy array
transformed features
- Return type
features
-
transform
(self, features)¶ Transforms features.
- Parameters
features (numpy array) – input features
- Returns
- numpy array
transformed features
- Return type
features
-
compute_dimension_reduction_params
(self, data_dim)¶ Compute dimensions reduction parameters.
- Parameters
data_dim (int) – dimension of the data
-
apply_padding
(self, features)¶ Apply padding to data
- Parameters
features (numpy array) – features to be padded
-