PODAE

Module for FNN-Autoencoders.

class PODAE(pod, ae)[source]

Bases: POD, AE

Combined POD and AutoEncoder reduction class.

This class first applies POD to reduce the dimensionality, then uses an autoencoder for further reduction in the latent space.

Parameters:

  • pod (POD) – The POD instance for initial reduction.

  • ae (AE) – The AutoEncoder instance for latent space reduction.

Initialize the PODAE reducer.

Parameters:

  • pod (POD) – The POD instance.

  • ae (AE) – The AutoEncoder instance.

_abc_impl = <_abc._abc_data object>
expand(g)[source]

Projects a reduced to full order solution.

Param:

numpy.ndarray g the latent variables.

Note

Same as inverse_transform. Kept for backward compatibility.

fit(X)[source]

Fit the PODAE on the snapshots.

First applies POD, then trains the autoencoder on POD coefficients.

Parameters:

X (numpy.ndarray) – The input snapshots matrix (stored by column).

inverse_transform(g)[source]

Projects a reduced to full order solution.

Param:

numpy.ndarray g the latent variables.

reduce(X)[source]

Reduces the given snapshots.

Parameters:

X (numpy.ndarray) – the input snapshots matrix (stored by column).

Note

Same as transform. Kept for backward compatibility.

transform(X)[source]

Reduces the given snapshots.

Parameters:

X (numpy.ndarray) – the input snapshots matrix (stored by column).