Class Dictionary
A Dictionary is a vector-valued function denoted as
\[\Psi = (\psi_1, \psi_2, \cdots, \psi_{N_\psi})^T,\]
where each component \(\psi_i = g_i\) for \(i = 1,\cdots,N_y\). The functions \(g_{1},\cdots, g_{N_{y}}\) represent the target observable functions. It supports batched operation \(\Psi: \mathbb{R}^{N \times N_x} \rightarrow \mathbb{R}^{N \times N_{\psi}}\).
API Documentation¶
dim_input
property
¶
Input dimension.
Returns:
| Type | Description |
|---|---|
int
|
The input dimension. |
dim_output
property
¶
Output dimension.
Returns:
| Type | Description |
|---|---|
int
|
The output dimension. |
__call__(x)
¶
Apply the dictionary to the input.
Parameters:
| Name | Type | Description | Default |
|---|---|---|---|
x
|
Tensor
|
Input state of the size \((N, N_x)\) |
required |
Returns:
| Type | Description |
|---|---|
Tensor
|
The basis functions of the size \((N, N_\psi)\) |
__init__(function, dim_input, dim_output)
¶
Dictionary \(\Psi\).
Parameters:
| Name | Type | Description | Default |
|---|---|---|---|
function
|
torch.Tensor -> torch.Tensor
|
Dictionary basis function \((N, N_x) \rightarrow (N, N_{\psi})\) |
required |
dim_input
|
int
|
Input dimension \(N_x\). |
required |
dim_output
|
int
|
Output dimension \(N_\psi\). |
required |