Skip to content

Commit 7d55010

Browse files
authored
Merge pull request #2223 from pymc-devs/bump_v3.1rc4
Bump version to 3.1rc4
2 parents 901cb20 + fe7ac3e commit 7d55010

28 files changed

+2410
-3936
lines changed

CHANGELOG.md

Lines changed: 0 additions & 13 deletions
This file was deleted.

RELEASE-NOTES.md

Lines changed: 31 additions & 7 deletions
Original file line numberDiff line numberDiff line change
@@ -1,13 +1,29 @@
11
# Release Notes
22

3-
## PyMC3 3.1 (TBA)
3+
## PyMC3 3.1 (June 23, 2017)
44

55
### New features
66

7-
* [Add Gaussian Process submodule](http://pymc-devs.github.io/pymc3/notebooks/GP-introduction.html)
7+
* New user forum at http://discourse.pymc.io
8+
9+
* [Gaussian Process submodule](http://pymc-devs.github.io/pymc3/notebooks/GP-introduction.html)
10+
11+
* Much improved variational inference support:
12+
13+
- [Add Operator Variational Inference (experimental).](http://pymc-devs.github.io/pymc3/notebooks/bayesian_neural_network_opvi-advi.html)
14+
15+
- [Add Stein-Variational Gradient Descent as well as Amortized SVGD (experimental).](https://github.com/pymc-devs/pymc3/pull/2183)
16+
17+
- [Add pm.Minibatch() to easily specify mini-batches.](http://pymc-devs.github.io/pymc3/notebooks/bayesian_neural_network_opvi-advi.html#Minibatch-ADVI)
18+
19+
- Added various optimizers including ADAM.
20+
21+
- Stopping criterion implemented via callbacks.
822

923
* sample() defaults changed: tuning is enabled for the first 500 samples which are then discarded from the trace as burn-in.
1024

25+
* MvNormal supports Cholesky Decomposition now for increased speed and numerical stability.
26+
1127
* Many optimizations and speed-ups.
1228

1329
* NUTS implementation now matches current Stan implementation.
@@ -28,24 +44,32 @@
2844

2945
* Much improved variational inference support:
3046

31-
- [Add Operator Variational Inference (experimental).](http://pymc-devs.github.io/pymc3/notebooks/bayesian_neural_network_opvi-advi.html)
47+
- [Added Operator Variational Inference (experimental).](http://pymc-devs.github.io/pymc3/notebooks/variational_api_quickstart.html)
3248

33-
- [Add Stein-Variational Gradient Descent as well as Amortized SVGD (experimental).](https://github.com/pymc-devs/pymc3/pull/2183)
49+
- [Added Stein-Variational Gradient Descent as well as Amortized SVGD (experimental).](https://github.com/pymc-devs/pymc3/pull/2183)
3450

35-
- [Add pm.generator() to easily specify mini-batches.](http://pymc-devs.github.io/pymc3/notebooks/bayesian_neural_network_opvi-advi.html#Minibatch-ADVI)
51+
- [Added `Minibatch` to easily specify mini-batches.](http://pymc-devs.github.io/pymc3/notebooks/bayesian_neural_network_opvi-advi.html#Minibatch-ADVI)
52+
53+
- Added full-rank ADVI
3654

3755
- Added various optimizers including ADAM.
56+
57+
- Deprecated old ADVI interface
58+
59+
- implemented `fit` function as the primary interface to approximation algorithms
60+
61+
* Added support for multidimensional minibatches
3862

3963
* [Sampled posteriors can now be turned into priors for Bayesian updating with a new interpolated distribution.](https://github.com/pymc-devs/pymc3/pull/2163)
4064

65+
* Added `Approximation` class and the ability to convert a sampled trace into an approximation via its `Empirical` subclass.
66+
4167
* `Model` can now be inherited from and act as a base class for user specified models (see pymc3.models.linear).
4268

4369
* Add MvGaussianRandomWalk and MvStudentTRandomWalk distributions.
4470

4571
* GLM models do not need a left-hand variable anymore.
4672

47-
* Add support for cholesky parametrizations for cov/corr matrices.
48-
4973
* Refactored HMC and NUTS for better readability.
5074

5175
* Add support for Python 3.6.

docs/source/conf.py

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -78,7 +78,7 @@
7878
# The short X.Y version.
7979
version = '3.1'
8080
# The full version, including alpha/beta/rc tags.
81-
release = '3.1rc3'
81+
release = '3.1rc4'
8282

8383
# The language for content autogenerated by Sphinx. Refer to documentation
8484
# for a list of supported languages.

docs/source/notebooks/Diagnosing_biased_Inference_with_Divergences.ipynb

Lines changed: 298 additions & 170 deletions
Large diffs are not rendered by default.

docs/source/notebooks/GLM-hierarchical-advi-minibatch.ipynb

Lines changed: 69 additions & 43 deletions
Large diffs are not rendered by default.

docs/source/notebooks/GP-introduction.ipynb

Lines changed: 0 additions & 673 deletions
This file was deleted.

docs/source/notebooks/GP-slice-sampling.ipynb

Lines changed: 24 additions & 38 deletions
Large diffs are not rendered by default.

docs/source/notebooks/GP-smoothing.ipynb

Lines changed: 18 additions & 71 deletions
Large diffs are not rendered by default.

docs/source/notebooks/LKJ.ipynb

Lines changed: 35 additions & 154 deletions
Large diffs are not rendered by default.

docs/source/notebooks/Model Comparison.ipynb

Lines changed: 116 additions & 114 deletions
Large diffs are not rendered by default.

0 commit comments

Comments
 (0)