Scikit-ReduceModel’s documentation!

Scikit-ReduceModel is a Python package to construct reduced model. This code is an extension of the standard reduced-base framework and provides an efficient and accurate solution for model building. It implements the hp-greedy refinement strategy, an enhancement approach for reduced-base model building. The approach uses a parameter space partitioning, a local reduced basis and a binary tree as the resulting structure, all obtained automatically.

Contents:

• Installation
• Tutorials
  • Tutorial | Simple pendulum
  • Tutorial | Gravitational wave science
• skreducedmodel
  • skreducedmodel package
• License

Indices and tables

• Index

• Module Index

• Search Page

Quick Usage

In order to construct a reduced model, we require knowledge of a training set (training_set). That is, we need to be familiar with a set of functions parameterized by a real number λ, denoted as \(f_λ(x)\).

We need also a distretization of the \(x\) (x_set) and of the \(λ\) space (param).

Then, we can first built the reduced basis, in this case, we use the default parameters.

from skreducedmodel.reducedbasis import ReducedBasis

rb = ReducedBasis()
rb.fit(training_set = training_set,
      parameters = param
      physical_points = x_set)

The second step is built the empirical interpolator with the reduced basis generated

from skreducedmodel.empiricalinterpolation import EmpiricalInterpolation

eim = EmpiricalInterpolation(rb)
eim.fit()

Finally, we construct the reduced model from our eim object

from skreducedmodel.surrogate import Surrogate

model = Surrogate(eim)
model.fit()

In case we are interested in studying the ReducedBasis and EmpiricalInterpolation objects, the package has a function that automates the whole process.

from skreducedmodel.mksurrogate import mksurrogate

surrogate = mksurrogate(parameters = param,
                        training_set = training_set,
                        physical_points = x_set,
                        )