POD Functions
These functions are used for the proper orthogonal decomposition of samples
calculate_v_and_betas(stacked_sub_displ, progress_bar=None, task_id=None)
Calculates the right reduced order Basis V and up to 10 eigenvalues of the subsamples
Parameters:
| Name | Type | Description | Default |
|---|---|---|---|
stacked_sub_displ |
np.ndarray
|
np.ndarray containing all subsampled displacements shape must be (samples, timesteps, nodes, dims) |
required |
Returns:
| Name | Type | Description |
|---|---|---|
v_big |
np.ndarray
|
Reduced order basis to transform betas bag into subsamples |
betas |
np.ndarray
|
Projected simulation runs |
err_msg |
str
|
Error message if not enough samples where provided |
svd_step_and_dim(s_mat, k=10)
Performs a svds operation on the two dimensional s_mat
Parameters:
| Name | Type | Description | Default |
|---|---|---|---|
s_mat |
np.ndarray
|
2D array on which the svds operation shall be performed |
required |
k |
The size of the POD |
10
|
Returns:
| Name | Type | Description |
|---|---|---|
v |
ndarray
|
Array containing the right reduced order basis |