Package mvpa :: Package clfs :: Module kernel :: Class KernelSquaredExponential
[hide private]
[frames] | no frames]

Class KernelSquaredExponential

source code


The Squared Exponential kernel class.

Note that it can handle a length scale for each dimension for Automtic Relevance Determination.

Instance Methods [hide private]
 
__init__(self, length_scale=1.0, sigma_f=1.0, **kwargs)
Initialize a Squared Exponential kernel instance.
source code
 
reset(self)
Resets the kernel dropping internal variables to the original values
source code
 
__repr__(self)
repr(x)
source code
 
compute(self, data1, data2=None)
Compute kernel matrix.
source code
 
set_hyperparameters(self, hyperparameter)
Set hyperaparmeters from a vector.
source code
 
compute_lml_gradient(self, alphaalphaT_Kinv, data)
Compute grandient of the kernel and return the portion of log marginal likelihood gradient due to the kernel. Shorter formula. Allows vector of lengthscales (ARD).
source code
 
compute_lml_gradient_logscale(self, alphaalphaT_Kinv, data)
Compute grandient of the kernel and return the portion of log marginal likelihood gradient due to the kernel. Hyperparameters are in log scale which is sometimes more stable. Shorter formula. Allows vector of lengthscales (ARD).
source code
 
_setlength_scale(self, v)
Set value of length_scale and its _orig
source code

Inherited from Kernel: compute_gradient

Inherited from object: __delattr__, __format__, __getattribute__, __hash__, __new__, __reduce__, __reduce_ex__, __setattr__, __sizeof__, __str__, __subclasshook__

Class Variables [hide private]
  length_scale = property(fget= lambda x: x._length_scale, fset=...
Properties [hide private]

Inherited from object: __class__

Method Details [hide private]

__init__(self, length_scale=1.0, sigma_f=1.0, **kwargs)
(Constructor)

source code 
Initialize a Squared Exponential kernel instance.
Parameters:
  • length_scale (float OR numpy.ndarray) - the characteristic length-scale (or length-scales) of the phenomenon under investigation. (Defaults to 1.0)
  • sigma_f (float) - Signal standard deviation. (Defaults to 1.0)
Overrides: object.__init__

reset(self)

source code 
Resets the kernel dropping internal variables to the original values
Overrides: Kernel.reset
(inherited documentation)

__repr__(self)
(Representation operator)

source code 
repr(x)
Overrides: object.__repr__
(inherited documentation)

compute(self, data1, data2=None)

source code 
Compute kernel matrix.
Parameters:
  • data1 (numpy.ndarray) - data
  • data2 (numpy.ndarray) - data (Defaults to None)
Overrides: Kernel.compute

set_hyperparameters(self, hyperparameter)

source code 

Set hyperaparmeters from a vector.

Used by model selection.

compute_lml_gradient(self, alphaalphaT_Kinv, data)

source code 
Compute grandient of the kernel and return the portion of log marginal likelihood gradient due to the kernel. Shorter formula. Allows vector of lengthscales (ARD).
Overrides: Kernel.compute_lml_gradient

compute_lml_gradient_logscale(self, alphaalphaT_Kinv, data)

source code 
Compute grandient of the kernel and return the portion of log marginal likelihood gradient due to the kernel. Hyperparameters are in log scale which is sometimes more stable. Shorter formula. Allows vector of lengthscales (ARD).
Overrides: Kernel.compute_lml_gradient_logscale

Class Variable Details [hide private]

length_scale

Value:
property(fget= lambda x: x._length_scale, fset= _setlength_scale)