stats package

Submodules

stats.kernels module

class stats.kernels.ExponentialQuadraticKernel(scale=None, lengthScale=None, dtype=torch.float64)[source]

Bases: stats.kernels.Kernel

buildKernelMatrix(X1, X2=None)[source]
buildKernelMatrixDiag(X)[source]
class stats.kernels.Kernel[source]

Bases: abc.ABC

abstract buildKernelMatrix(X1, X2=None)[source]
abstract buildKernelMatrixDiag(X)[source]
getParams()[source]
setParams(params)[source]
class stats.kernels.PeriodicKernel(scale=None, lengthScale=None, period=None, dtype=torch.float64)[source]

Bases: stats.kernels.Kernel

buildKernelMatrix(X1, X2=None)[source]
buildKernelMatrixDiag(X)[source]

stats.sampler module

class stats.sampler.Sampler[source]

Bases: object

sampleInhomogeneousPP_thinning(intensityFun, T, dt=0.03)[source]

Thining algorithm to sample from an inhomogeneous point process. Algorithm 2 from Yuanda Chen (2016). Thinning algorithms for simulating Point Prcesses.

intensityFun: function

Intensity function of the point process.

T: double

The returned samples of the point process will be in [0, T]

nGrid: integer

number of points in the grid used to search for the maximum of intensityFun.

inhomogeneous: list

samples of the inhomogeneous point process with intensity function intensityFun.

homogeneous: list

samples of the homogeneous that was filtered to generate the inhomogeneous point process.

sampleInhomogeneousPP_timeRescaling(intensityFun, T, dt=0.03)[source]

Time rescaling algorithm to sample from an inhomogeneous point process. Chapter 2 from Uri Eden papers/numericalMethods/uri-eden-point-process-notes.pdf

intensityFun: function

Intensity function of the point process.

T: double

The returned samples of the point process will be in [0, T]

nGrid: integer

number of points in the grid used to search for spike times.

list

samples of the inhomogeneous point process with intensity function intensityFun.

Module contents