michaelhush/M-LOOP

When optimizing one parameter, there were some issues reimporting the
saved files for the visualizations to work. This was due to the
problematic corner case of zero D or one D with one element arrays in
numpy. This has now been sanitized. Also fixed some critical typos in
the documentation.

Fixed a bug where an attribute wasn’t present in the learner class. Was
a problem when attempting to plot the visualizations from a file.

Added support for previous data files to be imported into a gaussian
process learner.

Adding NN from charmasaur

There appears to be some issues with multiprocessing and gaussian
process but only on MacOS, and possibly just my machine. So I’ve
removed all the testing statements I had in the previous commit.

Branch should be ready now to integrate in a genuine NN.

Visualizations now work for NN and GP learners.

Mysterious bug has appeared in GP. The scikit-learn stops providing
uncertainty predictions after being fit for a certain number of times.

Commiting so I can change branch and investigate.

Added a controller and learner for the neural net.

Also added a new class MachineLearnerController which GaussianProcess
and NeuralNet both inherit from.

I broke the visualizations for GPs in this update. But all the tests
work.

Added some updates to docstrings and test unit parameters.

Updated the documentation.
Candidate for new version to be released on PyPI

Fixed halting conditions and bad flags

Previously the training runs had to be completed before M-LOOP would
halt. This lead to unintuitive behavior when the halting conditions
were early on in the optimization process.

M-LOOP now halts immediately when any of the halting conditions are
met.

Documentation and setup tweaks

Candidate for version 2.1.0 release.

Docupdate

Update to documentation complete. Now describes how to use the shell
interface, the differential evolution optimizer and using M-LOOP as an
MPI.

Documentation has been updated to explain all the added features and
also how to use M-LOOP as a python API. Still needs proof reading.

Updating the documentation to prepare for the 2.1.0 release.

Deleted unnecessary line from TravisCI config.

Differential evolution added to M-LOOP.

Differential Evolution now been added to M-LOOP and is set to the
default trainer for the gaussian process. Tests and examples have been
added. The installation section of the documentation has also been
updated.

The differential evolution controller has been implemented and is now
set at the default trainer for gaussian process. Tests have been added
to the automated suite and there is some basic visualizations available.
Still have to debug a possible issue with GP andexamples logging and
extras.

First version of differential evolution complete. Next testing will be
investigated

Continuing the implementation.

Started modifying the differential evolution code from scipy for use in
M-LOOP.

Attempting to define the correct python version explicitly.

Copied syntax for conditional test from facebook example. Hopefully
will work now.

Added $ to variable name

Travis test 2.7 now works, but only after updating to 2.7.12 (why on
earth Travis CI defaults to 2.7.9 is beyond me. ) but trying to update
3.5 creates an error, attempted to add a conditional test to get around
this.

Setuptools updated

Travis CI is basically broken for python 2.7. Still trying to fix it.

The problem with Travis CI is it uses a very old version of pip when
testing on python 2.7. Attempted to add a command to upgrade pip first
before running installation.

Hopefully this fixes travis CI

Added requirements file and updated travis to make more conservative
tests.

Sciki-learn 0.18 has now officially been released as the new stable
version. It is also available in the anaconda distribution. We have
removed the scikit-learn code that was included with M-LOOP previously
and now have it has a dependency.

Command Line now works on windows as well as Linux and MacOS