Release 2.1.4 (GPflow#1613)

* Replace len(inducing_variable) with inducing_variable.num inducing property (GPflow#1594).

  Adds support for inducing variables with dynamically changing shape. Change usage from `len(inducing_variable)` to `inducing_variable.num_inducing` instead. Resolves GPflow#1578.

* HeteroskedasticTFPConditional should construct tensors at class-construction, not at module-import time (GPflow#1598)

Merge pull request GPflow#1593 from GPflow/develop

Release 2.1.3

Merge pull request GPflow#1576 from GPflow/develop

Release 2.1.2

Merge pull request GPflow#1565 from GPflow/develop

Release 2.1.1 develop -> master

Release 2.1.0 (GPflow#1552)

* Update pull request template (GPflow#1510)

Clarify template to make it easier for contributors to fill in relevant information.

* Temporary workaround for tensorflow_probability dependency issue (GPflow#1522)

* pin cloudpickle==1.3.0 as temporary workaround for tensorflow/probability#991 to unblock our build (to be reverted once fixed upstream)

* Update readme with new project using GPflow (GPflow#1530)

* fix bug in varying_noise notebook (GPflow#1526)

* Fix formatting in docs (intro.md) and restore link removed by GPflow#1498 (GPflow#1520)

* pin tensorflow<2.3 tensorflow-probability<0.11 (GPflow#1537)

* Quadrature Refactoring (GPflow#1505)

* WIP: quadrature refactoring

* Removing old ndiagquad code

* deleted test code

* formatting and type-hint

* merge modules

* black formatting

* formatting

* solving failing tests

* fixing failing tests

* fixes

* adapting tests for new syntax, keeping numerical behavior

* black formatting

* remove printf

* changed code for compiled tf compatibility

* black

* restored to original version

* undoing changes

* renaming

* renaming

* renaming

* reshape kwargs

* quadrature along axis=-2, simplified broadcasting

* black

* docs

* docs

* helper function

* docstrings and typing

* added new and old quadrature equivalence tests

* black

* Removing comments

Co-authored-by: Vincent Dutordoir <[email protected]>

* Typo

Co-authored-by: Vincent Dutordoir <[email protected]>

* notation

Co-authored-by: Vincent Dutordoir <[email protected]>

* reshape_Z_dZ return docstring fix

* FIX: quad_old computed with the ndiagquad_old

Co-authored-by: Vincent Dutordoir <[email protected]>

* more readable implementation

Co-authored-by: Vincent Dutordoir <[email protected]>

* tf.ensure_shape added

* removed ndiagquad

* removed ndiagquad

* Revert "removed ndiagquad"

This reverts commit 7bb0e9f.

* FIX: shape checking of dZ

* Revert "removed ndiagquad"

This reverts commit 8e23524.

Co-authored-by: Gustavo Carvalho <[email protected]>
Co-authored-by: ST John <[email protected]>
Co-authored-by: Vincent Dutordoir <[email protected]>

* Add base_conditional_with_lm function (GPflow#1528)

* Added base_conditional_with_lm function, which accepts Lm instead of Kmm

Co-authored-by: Neil Ferguson <[email protected]>
Co-authored-by: Vincent Dutordoir <[email protected]>
Co-authored-by: st-- <[email protected]>

* Fixed separate_independent_conditional to correctly handle q_sqrt=None. (GPflow#1533)

* Fixed separate_independent_conditional to correctly handle q_sqrt=None.

Co-authored-by: Aidan Scannell <[email protected]>
Co-authored-by: st-- <[email protected]>

* Bump version numbers to 2.1.0. (GPflow#1544)

* Re-introduce pytest-xdist (GPflow#1541)

Enables pytest-xdist for locally running tests (`make test`) on multiple cores in parallel.

* check dependency versions are valid on CI (GPflow#1536)

* Update to not use custom image (GPflow#1545)

* Update to not use custom image

* Add test requirements

* Update parameter to be savable (GPflow#1518)

* Fix for quadrature failure mode when autograph was set to False (GPflow#1548)

* Fix and test

* Change shape of quadrature tensors for better broadcasting (GPflow#1542)

* using the first dimension to hold the quadrature summation

* adapting ndiagquad wrapper

* Changed bf for bX in docstrings

Co-authored-by: Gustavo Carvalho <[email protected]>
Co-authored-by: st-- <[email protected]>
Co-authored-by: Vincent Dutordoir <[email protected]>

* Update min TFP supported version to 0.10 (GPflow#1551)

* Broadcasting constant and zero mean function (GPflow#1550)

* Broadcasting constant and zero mean function

* Use rank instead of ndim

Co-authored-by: st-- <[email protected]>
Co-authored-by: joelberkeley-pio <[email protected]>
Co-authored-by: gustavocmv <[email protected]>
Co-authored-by: Gustavo Carvalho <[email protected]>
Co-authored-by: ST John <[email protected]>
Co-authored-by: Neil Ferguson <[email protected]>
Co-authored-by: Neil Ferguson <[email protected]>
Co-authored-by: Aidan Scannell <[email protected]>
Co-authored-by: Aidan Scannell <[email protected]>
Co-authored-by: Sandeep Tailor <[email protected]>
Co-authored-by: Artem Artemev <[email protected]>

Merge pull request GPflow#1506 from GPflow/develop

Release 2.0.5

Merge pull request GPflow#1481 from GPflow/develop

Release 2.0.4

Merge branch 'develop'

Release 2.0.2 (GPflow#1459)

* add long_description and project_urls to setup.py (GPflow#1438)
* add type hints for probability distributions (GPflow#1421)
* remove comment (GPflow#1441)
* Use a type alias in the fnction signature of leading_transpose (GPflow#1442)
* refactor natgrads to be more efficient (GPflow#1443)
* Fix dimensions of kernel evaluation of changepoint kernel (GPflow#1446)
* Removed unusued imports. (GPflow#1450)
* Improve representation of GPflow objects in IPython/Jupyter notebook (GPflow#1453)
* includes the repr() string in IPython/Jupyter notebook representation as well (i.e. fully-qualified class name and object hash (memory address), which helps distinguish objects from each other)
* only displays the parameter table when it is not empty
* makes use of default_summary_fmt() for IPython shell
* Convert data structures to tensor in model init method (GPflow#1452)
* Use a boolean for full covariance in sample_mvn. (GPflow#1448)
* GPflow#1452 for GPMC model (GPflow#1458)
* release candidate v2.0.2 (GPflow#1457)

Co-authored-by: st-- <[email protected]>
Co-authored-by: joelberkeley-pio <[email protected]>
Co-authored-by: John Mcleod <[email protected]>
Co-authored-by: Mark van der Wilk <[email protected]>
Co-authored-by: Artem Artemev <[email protected]>

Release 2.0.1 (GPflow#1436)

Release notes:
- Improve structure of likelihoods subdirectory (GPflow#1416)
- Update README.md (GPflow#1401) and GPflow 2 upgrade guide (GPflow#1414)
- Improved handling of invalid values for constrained Parameters (GPflow#1408)
- Improvements on types/function annotations (GPflow#1406, GPflow#1420)
- Documentation improvements (metalearning with GPs: GPflow#1382, coregionalization notebook: GPflow#1402, MCMC notebook: GPflow#1410, intro to gpflow with tensorflow 2: GPflow#1413)
- Minor documentation fixes (GPflow#1429, GPflow#1430, GPflow#1433)
- Fix: move matplotlib import inside ImageToTensorBoard (GPflow#1399)
- Fix: tf.function compilation of ndiagquad (GPflow#1418)
- Fix: cache tensorboard file writers and re-use them (GPflow#1424)