Kernels (anglepy.kernels)

This module contains the kernel implementations to make the geodesic energy score a strictly proper scoring rule. For more details, refer to Table 1 of https://projecteuclid.org/journalArticle/Download?urlId=10.3150%2F12-BEJSP06.

anglepy.kernels.apply_kernel(d, kernel_func)

Unified function. Returns the negative of the provided kernel function.

anglepy.kernels.askey(c=1, tau=2)

c > 0, tau >= 2

anglepy.kernels.c2_wendland(c=3.141592653589793, tau=4)

0 < c <= pi, tau >= 4

anglepy.kernels.c4_wendland(c=3.141592653589793, tau=6)

0 < c <= pi, tau >= 6

anglepy.kernels.dagum(c=1, tau=1, alpha=0.5)

c > 0, 0 < tau <= 1, 0 < alpha < tau

anglepy.kernels.generalized_cauchy(c=1, tau=1, alpha=1)

c > 0, tau > 0, 0 < alpha <= 1

anglepy.kernels.multiquadric(tau=1, delta=0.5)

tau > 0, 0 < delta < 1

anglepy.kernels.powered_exponential(c=1, alpha=1)

c > 0, 0 < alpha <= 1

anglepy.kernels.sine_power(alpha=1)

0 < alpha < 2

anglepy.kernels.spherical(c=1)

c > 0