ContinuousModel#
- class ContinuousModel(operators, solver=None)[source]#
Nonparametric system of ordinary differential equations \(\ddt\qhat(t) = \fhat(\qhat(t), \u(t))\).
Here,
\(\qhat(t)\in\RR^{r}\) is the model state, and
\(\u(t)\in\RR^{m}\) is the (optional) input.
The structure of \(\fhat\) is specified through the
operators
argument.- Parameters:
- operatorslist of
opinf.operators
objects Operators comprising the terms of the model.
- solver
opinf.lstsq
object or float > 0 or None Solver for the least-squares regression. Defaults:
None
:opinf.lstsq.PlainSolver
. SVD-based solve without regularization.float > 0:
opinf.lstsq.L2Solver
. SVD-based solve with scalar Tikhonov regularization.
- operatorslist of
Properties:- A_#
opinf.operators.LinearOperator
(orNone
).
- B_#
opinf.operators.InputOperator
(orNone
).
- G_#
opinf.operators.CubicOperator
(orNone
).
- H_#
opinf.operators.QuadraticOperator
(orNone
).
- N_#
opinf.operators.StateInputOperator
(orNone
).
- c_#
opinf.operators.ConstantOperator
(orNone
).
- input_dimension#
Dimension \(m\) of the input (zero if there are no inputs).
- operator_matrix#
\(r \times d(r, m)\) operator matrix, e.g., \(\Ohat = [~\chat~~\Ahat~~\Hhat~~\Bhat~]\).
This matrix does not includes the entries of any operators whose entries are known a priori.
- operators#
Operators comprising the terms of the model.
- solver#
Solver for the least-squares regression, see
opinf.lstsq
.
- state_dimension#
Dimension \(r\) of the state.
Methods:Make a copy of the model.
Learn the model operators from data.
Construct a reduced-order model by taking the (Petrov-)Galerkin projection of each model operator.
Sum the state Jacobian of each model operator.
Load a serialized model from an HDF5 file, created previously from the
save()
method.Solve the system of ordinary differential equations.
Solve the Operator Inference regression using the data from the last
fit()
call, then extract the inferred operators.Evaluate the right-hand side of the model by applying each operator and summing the results.
Serialize the model, saving it in HDF5 format.