TransformerMulti

class TransformerMulti(transformers, variable_sizes=None)

Join transformers together for states with multiple variables.

This class is for states that can be written (after discretization) as

\[\begin{split}\q = \left[\begin{array}{c} \q_{0} \\ \q_{1} \\ \vdots \\ \q_{n_q - 1} \end{array}\right] \in \RR^{n},\end{split}\]

where each \(\q_{i} \in \RR^{n_i}\) represents a single discretized state variable of dimension \(n_i \in \NN\). The dimension of the joint state \(\q\) is \(n = \sum_{i=0}^{n_q - 1} n_i\), i.e., state_dimension = sum(variable_sizes). Individual transformers are calibrated for each state variable.

Parameters:

transformerslist

Initialized (but not necessarily trained) transformer objects, one for each state variable. Entries of this list can be None, in which case a NullTransformer is used.

variable_sizeslist or None

Dimensions for each state variable, \(n_0,\ldots,n_{n_q-1}\). If None (default), set \(n_i\) to transformers[i].state_dimension; if any of these are None, assume all state variables have the same dimension, i.e., \(n_0 = n_1 = \cdots = n_x \in \NN\) with \(n_x\) to be determined in fit(). In this case, \(n = n_q n_x\).

Notes

This class connects multiple transformers “vertically” by assigning different transformations for different parts of the state; see TransformerPipeline to connect multiple transformers “horizontally”, i.e., to perform one transformation after another on the entire state.

Properties:

num_variables

Number of state variables \(n_q \in \NN\).

state_dimension

Total dimension \(n = \sum_{i=0}^{n_q-1} n_i \in \NN\) of the joint state.

transformers

Transformers for each state variable.

variable_names

Names for each state variable.

variable_sizes

Dimensions of each state variable, \(n_0,\ldots,n_{n_q-1}\in\NN\).

Methods:

fit	Learn (but do not apply) the transformation.
fit_transform	Learn and apply the transformation.
get_var	Extract a single variable from the joint state.
inverse_transform	Apply the inverse of the learned transformation.
load	Load a previously saved transformer from an HDF5 file.
save
split	Split the joint state into the individual state variables.
transform	Apply the learned transformation.
transform_ddts	Apply the learned transformation to snapshot time derivatives.
verify	Verify that transform() and inverse_transform() are consistent and that transform_ddts(), if implemented in each transformer, is consistent with transform().