Kernels

Kernels are used to define trends in distribution parameters with regards to specific covariates. They can be as complex as necessary but we provide by default a set of common kernels that can be used directly or as a base for more complex ones.

Class

Kernel(f[, x, fname])	Represents a kernel function of one covariate with parameters.
constant([value])	A kernel representing a constant value.

Methods

with_covariate(covariate)

Create a new instance of the kernel with the given covariate.

Functions

linear(X, a, b)	Linear kernel function.
polynomial(X, a, b, c)	Polynomial kernel function.
exponential(X, a, b)	Exponential kernel function.
exponential_ratio(X, a, b, c)	Exponential ratio kernel function.
trigonometric(X, a, b, c)	Trigonometric kernel function.
linear_regression(x[, add_intercept])	Linear regression of the columns in the data.
exponential_linear_regression(x[, add_intercept])	Exponential of a linear sum of the columns in the data.
polynomial_regression(x[, degree])	Polynomial regression in the columns of the data.
categories_qualitative(x[, fixed_values])	Kernel for qualitative (categorical) data.
hawkes(X, mu, alpha, theta)	Hawkes process with exponential kernel.