Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / NICTA/revrand issues and pull requests

#133 - Publish Nlopt wrappers of old revrand project as separate package

Issue - State: open - Opened by dschmitz89 over 3 years ago - 4 comments

#132 - How to Simulate A la Carte

Issue - State: open - Opened by FBoyang over 4 years ago - 2 comments

#131 - problems with demos

Issue - State: closed - Opened by AIAdventures over 7 years ago - 4 comments

#130 - Make ARD length-scale bases return generator gradients instead of tensors

Issue - State: open - Opened by dsteinberg over 7 years ago
Labels: enhancement

#129 - Use scikit learn "check_estimator" in tests

Issue - State: closed - Opened by dsteinberg over 7 years ago - 1 comment

#128 - Implement random arc-cosine features

Issue - State: closed - Opened by dsteinberg almost 8 years ago - 1 comment
Labels: enhancement

#127 - Explore softplus as a replacement to log for optimising positive parameters

Issue - State: closed - Opened by dsteinberg almost 8 years ago - 1 comment
Labels: question

#126 - Random/GridSearchCv tests

Pull Request - State: closed - Opened by basaks almost 8 years ago

#125 - Grid/RandomSearchCv tests

Pull Request - State: closed - Opened by basaks almost 8 years ago

#124 - Check uncoverml revrand gridsearch traceback

Issue - State: closed - Opened by basaks almost 8 years ago - 2 comments

#123 - Use slices in calculating composed basis gradients rather than zero padding?

Issue - State: open - Opened by dsteinberg almost 8 years ago
Labels: enhancement

#122 - Make nice __repr__ methods for all of the objects in revrand

Issue - State: closed - Opened by dsteinberg almost 8 years ago

#119 - Basis functions should have "gain" parameters

Issue - State: closed - Opened by dsteinberg almost 8 years ago - 1 comment
Labels: enhancement

#118 - Trivial item, base.py comment unclear meaning

Issue - State: closed - Opened by zunzun almost 8 years ago - 1 comment
Labels: in progress

#117 - Should the GLM and SLM set RandomState of input objects?

Issue - State: closed - Opened by dsteinberg almost 8 years ago - 1 comment
Labels: question

#116 - Investigate numerical issues in tests

Issue - State: closed - Opened by dsteinberg almost 8 years ago - 1 comment
Labels: bug

#114 - Add "check_random_state" functionality to the random bases

Issue - State: closed - Opened by dsteinberg almost 8 years ago
Labels: enhancement

#113 - Import a few more revrand classes/functions into the main namespace?

Issue - State: closed - Opened by dsteinberg about 8 years ago
Labels: question

#112 - remove/archive legacyGP?

Issue - State: closed - Opened by dsteinberg about 8 years ago
Labels: question

#111 - Non-intrusive metrics logger

Pull Request - State: closed - Opened by darrnshn about 8 years ago - 2 comments

#110 - Numerical overflow in logtrick

Issue - State: closed - Opened by dsteinberg about 8 years ago - 2 comments
Labels: bug, in progress

#109 - combination basis error with GLM

Issue - State: closed - Opened by MushroomHunting about 8 years ago - 3 comments
Labels: bug

#108 - Implement Auto-Encoding VB for the GLM

Issue - State: closed - Opened by dsteinberg about 8 years ago
Labels: enhancement

#107 - Initialise the weights of the GLM with the MAP solution

Issue - State: closed - Opened by dsteinberg about 8 years ago
Labels: enhancement

#106 - Extensive profiling

Issue - State: open - Opened by dsteinberg about 8 years ago
Labels: testing

#105 - Implement predict_proba methods for the SLM and GLM

Issue - State: closed - Opened by dsteinberg about 8 years ago - 3 comments
Labels: enhancement

#104 - Finish report

Issue - State: closed - Opened by dsteinberg about 8 years ago
Labels: in progress

#103 - Do we need to use append_or_extend with the optimisers?

Issue - State: closed - Opened by dsteinberg about 8 years ago
Labels: question

#102 - Improve Optimisation and Model tests

Issue - State: closed - Opened by dsteinberg about 8 years ago - 1 comment
Labels: testing

#101 - Convert all demos to notebooks

Issue - State: closed - Opened by dsteinberg about 8 years ago
Labels: enhancement

#100 - We are still getting Non PSD errors in solve_posdef

Issue - State: closed - Opened by dsteinberg about 8 years ago - 1 comment
Labels: bug

#99 - Simple global optimisation for SLM (and GLM)

Issue - State: closed - Opened by dsteinberg about 8 years ago
Labels: enhancement, in progress

#98 - Variance of prediction results with BLR and GLM (and a few questions)

Issue - State: closed - Opened by MushroomHunting about 8 years ago - 3 comments

#97 - Return named tuple from glm and slm learn functions

Issue - State: closed - Opened by dsteinberg over 8 years ago - 2 comments
Labels: enhancement

#96 - Sampling in GLM predict functions has slowed down

Issue - State: closed - Opened by dsteinberg over 8 years ago
Labels: bug

#95 - Implement three parameter beta likelihood

Issue - State: closed - Opened by dsteinberg over 8 years ago
Labels: enhancement

#94 - Implement predict_likelihood in glm

Issue - State: closed - Opened by dsteinberg over 8 years ago
Labels: enhancement

#93 - Gaussian Spectral Mixture using RFF approximation

Issue - State: open - Opened by dsteinberg over 8 years ago
Labels: enhancement

#92 - Improve documentation

Issue - State: closed - Opened by dsteinberg over 8 years ago - 1 comment
Labels: enhancement

#91 - Feature/clean

Pull Request - State: closed - Opened by ltiao over 8 years ago

#90 - Cannot install :(

Issue - State: closed - Opened by ltiao over 8 years ago - 7 comments

#89 - Code review 10 June

Pull Request - State: closed - Opened by dsteinberg over 8 years ago

#88 - Write up GLM and SLM in the report

Issue - State: closed - Opened by dsteinberg over 8 years ago - 1 comment

#87 - implement even more bases

Issue - State: open - Opened by dsteinberg over 8 years ago
Labels: enhancement

#86 - Allow per-datum likelihood arguments in GLM

Issue - State: closed - Opened by dsteinberg over 8 years ago
Labels: enhancement

#85 - Do we still need safe divide (safediv) with safechol?

Issue - State: closed - Opened by dsteinberg over 8 years ago - 1 comment
Labels: question

#84 - Replace jitchol with safechol

Issue - State: closed - Opened by dsteinberg over 8 years ago
Labels: enhancement

#83 - Instead of inputting initial basis paramaters in bparams lists, input them as part of basis function construction

Issue - State: closed - Opened by dsteinberg over 8 years ago - 2 comments
Labels: enhancement, question

#82 - Make the output of predict_meanvar etc more transparent

Issue - State: closed - Opened by dsteinberg over 8 years ago
Labels: enhancement

#81 - Replicate scipy/numpy heirarchy

Issue - State: closed - Opened by dsteinberg over 8 years ago
Labels: question

#80 - Reconsider logging and printing

Issue - State: closed - Opened by dsteinberg over 8 years ago - 1 comment
Labels: enhancement, question

#79 - __call__ members are not documented in sphinx

Issue - State: closed - Opened by dsteinberg over 8 years ago

#78 - Make the new bounds contain transformation information

Issue - State: closed - Opened by dsteinberg over 8 years ago - 2 comments
Labels: someday/maybe

#77 - Bring back demo_classification.py

Issue - State: closed - Opened by dsteinberg over 8 years ago
Labels: enhancement

#76 - Remove redundant code

Issue - State: closed - Opened by dsteinberg over 8 years ago

#75 - Rebase autograd branch

Issue - State: closed - Opened by ltiao over 8 years ago
Labels: someday/maybe

#74 - Investigate latest release of GPy and deprecate LegacyGP

Issue - State: closed - Opened by ltiao over 8 years ago - 1 comment
Labels: question

#73 - Investigate Autograd for GLM

Issue - State: closed - Opened by ltiao over 8 years ago
Labels: someday/maybe

#72 - Make revrand algorithm interface compatible with scikit learn

Issue - State: closed - Opened by dsteinberg over 8 years ago - 5 comments
Labels: enhancement, help wanted

#71 - Second order derivatives are being calculated when they should not

Issue - State: closed - Opened by dsteinberg over 8 years ago
Labels: bug, someday/maybe

#70 - Look into autograd with learn_sgd

Issue - State: closed - Opened by ltiao over 8 years ago - 2 comments

#69 - Fancy Unit Tests for Stochastic Optimization

Issue - State: open - Opened by ltiao over 8 years ago
Labels: someday/maybe

#68 - Do this milestone.

Issue - State: closed - Opened by ltiao over 8 years ago
Labels: help wanted

#67 - Function result type module and decorators

Issue - State: closed - Opened by ltiao over 8 years ago
Labels: someday/maybe

#66 - Make unit tests for the optimiser decorators that test the callable jac functions

Issue - State: closed - Opened by dsteinberg over 8 years ago
Labels: testing

#64 - Autograd ELBO where basis function has more than one hyperparameter

Issue - State: closed - Opened by ltiao over 8 years ago - 1 comment
Labels: someday/maybe

#63 - Utilize optimizer callbacks

Issue - State: closed - Opened by ltiao over 8 years ago - 1 comment

#62 - Delete obsolete branches

Issue - State: closed - Opened by ltiao over 8 years ago

#61 - Port existing code to use Autograd's wrapped scipy/numpy

Issue - State: closed - Opened by ltiao over 8 years ago
Labels: someday/maybe

#60 - Get free GP with Autograd mvn.logpdf

Issue - State: closed - Opened by ltiao over 8 years ago

#59 - Profile regression.learn with autograd

Issue - State: closed - Opened by ltiao over 8 years ago - 2 comments

#58 - Implement compose and concat for basis functions

Issue - State: closed - Opened by ltiao over 8 years ago

#57 - Port existing basis functions to new framework

Issue - State: closed - Opened by ltiao over 8 years ago

#55 - Add support for bounded types to Basis function framework

Issue - State: closed - Opened by ltiao over 8 years ago

#54 - Add support for bounded types in optimizer

Issue - State: closed - Opened by ltiao over 8 years ago

#53 - Implement superficial "slice" basis function

Issue - State: closed - Opened by ltiao over 8 years ago

#52 - Implement bounded parameter types

Issue - State: closed - Opened by ltiao over 8 years ago - 5 comments

#51 - Define primitive for `cho_solve`

Issue - State: closed - Opened by ltiao over 8 years ago
Labels: someday/maybe

#50 - Autograd with Basis framework

Issue - State: closed - Opened by ltiao over 8 years ago
Labels: someday/maybe

#49 - Investigate gradient operation on basis concatenation

Issue - State: closed - Opened by ltiao over 8 years ago - 1 comment

#48 - Reimplement regression.learn with auto differentiated gradients

Issue - State: closed - Opened by ltiao over 8 years ago - 1 comment

#47 - Compare auto gradient values agains explicit gradients

Issue - State: closed - Opened by ltiao over 8 years ago - 3 comments

#46 - Slight API change to legacygp

Pull Request - State: closed - Opened by Kelvin-Hsu over 8 years ago - 1 comment

#45 - LinearBasis + RandomRBF_ARD breaks, hypers dimension inconsistent

Issue - State: closed - Opened by dsteinberg over 8 years ago - 1 comment
Labels: bug

#44 - dora now points to revrand/legacygp

Pull Request - State: closed - Opened by Kelvin-Hsu over 8 years ago - 1 comment

#43 - Prediction: in minibatches or per datum so to not blow out memory

Issue - State: closed - Opened by dsteinberg over 8 years ago
Labels: enhancement

#40 - LegacyGP gives me very small and negative predictive variances

Issue - State: closed - Opened by dsteinberg over 8 years ago
Labels: bug

#38 - Why do does regression.learn_sgd learn large variances?

Issue - State: closed - Opened by dsteinberg over 8 years ago - 3 comments
Labels: bug, question

#37 - Investigate automatic differentiation

Issue - State: closed - Opened by dsteinberg over 8 years ago - 1 comment
Labels: enhancement, question

#36 - Implement multitask regression

Issue - State: closed - Opened by dsteinberg almost 9 years ago - 2 comments
Labels: someday/maybe

#34 - Implement more basis functions

Issue - State: closed - Opened by dsteinberg almost 9 years ago
Labels: enhancement, help wanted

#33 - Get ADADELTA working on SPARK

Issue - State: closed - Opened by dsteinberg almost 9 years ago - 1 comment
Labels: someday/maybe

#7 - Applicative lifting of Basis functions a.k.a. "composing" (but not really composing)

Issue - State: closed - Opened by ltiao almost 9 years ago
Labels: enhancement

#4 - Optimizer abstraction

Issue - State: closed - Opened by ltiao almost 9 years ago - 2 comments
Labels: enhancement