Revision d1ac7b831ad36cd0e4bdd7980819f83208345148 authored by Alessandro Davide Ialongo on 07 February 2018, 13:36:57 UTC, committed by Mark van der Wilk on 07 February 2018, 14:00:17 UTC
* new implementations of RBF and Linear kernel expectations as well as the RBF-Linear cross kernel expectation * more efficient mean function expectation handling * Streamline quadrature expectations and unified the expectations files * clearer distinction between quadrature_expectation and expectation which makes @quadrature_fallback superfluous * "Private" _expectation methods are kept private * Improve handling of Sum kernels (with 1 bug fixed on when cross expectations simplify) * switch to assert_allclose instead of the less advisable assert_almost_equal * switch around (feat, kern) to (kern, feat) in expectations for consistency * improve test coverage and reduced error tolerance * add, correct and improve docstrings * make covariance name consistent for all Gaussian probability_distributions: p.cov * add shape inference for guassian covariances
1 parent da01a3f
roadmap.md
This document covers major planned development items for GPflow.
# Computational speed
- Add further benchmarks to [benchmark repository](https://github.com/GPflow/GPflowBenchmarks) including multiple GPUs.
- Incorporate remaining Tom Nickson GPU code from [branch](https://github.com/c0g/tomserflow) into TensorFlow main.
# Features
- Add ability to exploit Kronecker structure.
# Housekeeping
- See also issues marked "enhancement" in the GitHub [list](https://github.com/GPflow/GPflow/issues).
Computing file changes ...