mvpa2.clfs.gpr.GeneralizedLinearKernel

class mvpa2.clfs.gpr.GeneralizedLinearKernel(*args, **kwargs)

The linear kernel class.

Notes

Available conditional attributes:

• gradients: Dictionary of gradients per a parameter
• gradientslog: Dictionary of gradients per a parameter in logspace

(Conditional attributes enabled by default suffixed with +)

Methods

add_conversion(typename, methodfull, methodraw)	Adds methods to the Kernel class for new conversions
as_np()	Converts this kernel to a Numpy-based representation
as_raw_np()	Directly return this kernel as a numpy array.
cleanup()	Wipe out internal representation
compute(ds1[, ds2])	Generic computation of any kernel
computed(*args, **kwargs)	Compute kernel and return self
reset()

Initialize instance of GeneralizedLinearKernel

Parameters:

sigma_0 :

A simple constant squared value which is broadcasted across kernel. In the case of GPR – standard deviation of the Gaussian prior probability Normal(0, sigma_0**2) of the intercept of the linear regression. [Default: 1.0]

Sigma_p :

A generic scalar or vector, or diagonal matrix to scale all dimensions or associate different scaling to each dimensions while computing te kernel matrix: k(x_A,x_B) = x_A^\top \Sigma_p x_B + \sigma_0^2. In the case of GPR – a scalar or a diagonal of covariance matrix of the Gaussian prior probability Normal(0, Sigma_p) on the weights of the linear regression. [Default: 1.0]

enable_ca : None or list of str

Names of the conditional attributes which should be enabled in addition to the default ones

disable_ca : None or list of str

Names of the conditional attributes which should be disabled

Methods

add_conversion(typename, methodfull, methodraw)	Adds methods to the Kernel class for new conversions
as_np()	Converts this kernel to a Numpy-based representation
as_raw_np()	Directly return this kernel as a numpy array.
cleanup()	Wipe out internal representation
compute(ds1[, ds2])	Generic computation of any kernel
computed(*args, **kwargs)	Compute kernel and return self
reset()

reset()