ichor.core.models.mean package

Submodules

ichor.core.models.mean.constant module

class ConstantMean(value: float)

Bases: Mean

A constant value that is used for the mean function. When no training data is present close to test points, covariance is low, therefore the predictions for the test points will be this constant mean value.

Parameters:: value – A float to be used as the constant mean value

value(x: ndarray) → ndarray: Returns the constant mean value.

write_str() → str: Used to write the mean part of a model file

ichor.core.models.mean.linear module

class LinearMean(beta: ndarray, xmin: ndarray, ymin: float)

Bases: Mean

A linear mean.

value(x: ndarray) → ndarray: Returns the constant mean value.

write_str() → str: Used to write the mean part of a model file

ichor.core.models.mean.mean module

class Mean

Bases: ABC

Abstract base class for implementing different mean functions for a Gaussian Process.

abstract value(x: ndarray) → ndarray

abstract write_str() → str: Used to write the mean part of a model file

ichor.core.models.mean.quadratic module

class QuadraticMean(beta: ndarray, xmin: ndarray, ymin: float)

Bases: Mean

A quadratic mean

value(x: ndarray) → ndarray: Returns the constant mean value.

write_str() → str: Used to write the mean part of a model file

ichor.core.models.mean.zero module

class ZeroMean

Bases: Mean

Implements a zero mean for the Gaussian Process. When covariance between training points and a test point is low, this means the GP will return the mean of the GP (in this case it is set to 0).

value(x: ndarray) → ndarray: Return 0 as this is a Zero Mean Gaussian Process

write_str() → str: Used to write the mean part of a model file

Module contents

class ConstantMean(value: float)

Bases: Mean

A constant value that is used for the mean function. When no training data is present close to test points, covariance is low, therefore the predictions for the test points will be this constant mean value.

Parameters:: value – A float to be used as the constant mean value

value(x: ndarray) → ndarray: Returns the constant mean value.

write_str() → str: Used to write the mean part of a model file

class LinearMean(beta: ndarray, xmin: ndarray, ymin: float)

Bases: Mean

A linear mean.

value(x: ndarray) → ndarray: Returns the constant mean value.

write_str() → str: Used to write the mean part of a model file

class Mean

Bases: ABC

Abstract base class for implementing different mean functions for a Gaussian Process.

abstract value(x: ndarray) → ndarray

abstract write_str() → str: Used to write the mean part of a model file

class QuadraticMean(beta: ndarray, xmin: ndarray, ymin: float)

Bases: Mean

A quadratic mean

value(x: ndarray) → ndarray: Returns the constant mean value.

write_str() → str: Used to write the mean part of a model file

class ZeroMean

Bases: Mean

Implements a zero mean for the Gaussian Process. When covariance between training points and a test point is low, this means the GP will return the mean of the GP (in this case it is set to 0).

value(x: ndarray) → ndarray: Return 0 as this is a Zero Mean Gaussian Process

write_str() → str: Used to write the mean part of a model file