aim2dat.ml.kernels

Kernels to be used with scikit-learn models.

Functions

krr_ffprint_cosine(X, Y)

Cosine kernel for the F-Fingerprint Transformer based on the cosine distance metric

krr_ffprint_laplace(X, Y[, gamma])

Laplacian kernel for the F-Fingerprint Transformer based on the cosine distance metric

Module Contents

aim2dat.ml.kernels.krr_ffprint_cosine(X, Y)[source]

Cosine kernel for the F-Fingerprint Transformer based on the cosine distance metric (doi:10.1063/1.3079326).

Parameters:

  • X (numpy.array) – F-Fingerprint.

  • Y (numpy.array) – F-Fingerprint.

Returns:

float – Cosine kernel.

aim2dat.ml.kernels.krr_ffprint_laplace(X, Y, gamma=None)[source]

Laplacian kernel for the F-Fingerprint Transformer based on the cosine distance metric (doi:10.1063/1.3079326).

Parameters:

  • X (numpy.array) – F-Fingerprint.

  • Y (numpy.array) – F-Fingerprint.

Returns:

float – Laplacian kernel.