Class ParamKoopmanDLSolver
The class ParaKoopmanDLSolver implements the algorithm of
learning parametric Koopman decomposition.
API Documentation¶
__init__(dictionary)
¶
Initializes the KoopmanSolver with a given dictionary.
Parameters:
| Name | Type | Description | Default |
|---|---|---|---|
dictionary
|
TrainableDictionary
|
A dictionary containing configuration or parameters for the Koopman solver. |
required |
solve(dataset_train, dataset_val, paramkoopman, n_epochs, batch_size, tol=1e-06, lr_dic=0.0001, lr_koop=0.0001, weight=None)
¶
Solves the ParamKoopmanDL problem using the provided datasets and parameters.
Parameters:
| Name | Type | Description | Default |
|---|---|---|---|
dataset_train
|
KoopmanDataset
|
The training dataset. |
required |
dataset_val
|
KoopmanDataset
|
The validation dataset. |
required |
paramkoopman
|
ParamKoopman
|
The parameterized Koopman operator model. |
required |
n_epochs
|
int
|
The number of epochs to train the model. |
required |
batch_size
|
int
|
The batch size for training and validation. |
required |
tol
|
float
|
The tolerance for early stopping based on training loss. Default is 1e-6. |
1e-06
|
lr_dic
|
float
|
The learning rate for the dictionary optimizer. Default is 1e-4. |
0.0001
|
lr_koop
|
float
|
The learning rate for the Koopman optimizer. Default is 1e-4. |
0.0001
|
Returns:
| Type | Description |
|---|---|
ParamKoopman
|
The trained parameterized Koopman operator model. |