Skip to main content

emulators.linear

emulators.linear

Ordinary least squares (OLS) linear regression emulator.

The simplest emulator: fast to fit, but assumes a linear relationship between parameters and outputs and provides only limited uncertainty information. Useful for prototyping or for genuinely near-linear responses.

Classes

Name Description
LinearModel Emulator based on an ordinary least squares linear regression.
LinearModelScipy Emulator based on an ordinary least squares linear regression.

LinearModel

emulators.linear.LinearModel(x=None, y=None, test_fraction=0.25)

Emulator based on an ordinary least squares linear regression.

Fits a linear regression model that minimizes the residual sum of squares between the observed training targets and the targets predicted by the linear approximation. Fast, but assumes linearity and gives limited uncertainty estimates.

Example

import numpy as np, pandas as pd
from historymatching.emulators.linear import LinearModel
x = pd.DataFrame({'beta': np.random.rand(40), 'gamma': np.random.rand(40)})
y = pd.DataFrame({'peak': 2 * x['beta'] - x['gamma']})
em = LinearModel(x, y)
em.train()
pred = em.predict(x)

Methods

Name Description
get_hyperparameters Return linear model hyperparameters as a JSON-serializable dict.
predict Predict an output using the trained emulator.
print_emulator_description Display detailed specifications (for example, emulator coefficients)
train Fits a linear regression model to minimize the residual sum of squares between observed targets in the training data and the targets predicted by the linear approximation.
get_hyperparameters
emulators.linear.LinearModel.get_hyperparameters()

Return linear model hyperparameters as a JSON-serializable dict.

predict
emulators.linear.LinearModel.predict(x)

Predict an output using the trained emulator.

Parameters
Name Type Description Default
x pd.DataFrame Input data. Pandas dataframe with columns representing parameter values. required
Returns
Name Type Description
EmulationResults EmulationResults with predicted values and uncertainty intervals.
print_emulator_description
emulators.linear.LinearModel.print_emulator_description()

Display detailed specifications (for example, emulator coefficients) for the trained emulator.

train
emulators.linear.LinearModel.train()

Fits a linear regression model to minimize the residual sum of squares between observed targets in the training data and the targets predicted by the linear approximation.

LinearModelScipy

emulators.linear.LinearModelScipy(x=None, y=None, test_fraction=0.25)

Emulator based on an ordinary least squares linear regression. The emulator fits a linear regression model to minimize the residual sum of squares between observed targets in the training data and the targets predicted by the linear approximation.

Methods

Name Description
get_hyperparameters Return linear model (scipy) hyperparameters as a JSON-serializable dict.
predict Predict an output using the trained emulator.
print_emulator_description Display detailed specifications (for example, emulator coefficients)
train Fits a linear regression model to minimize the residual sum of squares between observed targets in the training data and the targets predicted by the linear approximation.
get_hyperparameters
emulators.linear.LinearModelScipy.get_hyperparameters()

Return linear model (scipy) hyperparameters as a JSON-serializable dict.

predict
emulators.linear.LinearModelScipy.predict(x)

Predict an output using the trained emulator.

Parameters
Name Type Description Default
x pd.DataFrame Input data. Pandas dataframe with columns representing parameter values. required
Returns
Name Type Description
EmulationResults EmulationResults with predicted values and uncertainty intervals.
print_emulator_description
emulators.linear.LinearModelScipy.print_emulator_description()

Display detailed specifications (for example, emulator coefficients) for the trained emulator.

train
emulators.linear.LinearModelScipy.train()

Fits a linear regression model to minimize the residual sum of squares between observed targets in the training data and the targets predicted by the linear approximation.