The basic idea is to have a metrics package, we can start with ROC/AUC (first on GPU, then if necessary on CPU). It should mimic the SKLearn API and results:

http://scikit-learn.org/stable/modules/generated/sklearn.metrics.roc_auc_score.html

Requirements:

1. GPU implementation. It can use Thrust and other libraries as first step.
2. Bindings for other languages (currently CTypes but soo

We should add BART as one final algorithm after xgboost and before SuperLearner - it's usually the best single model and it would be helpful to explain it to people.

Apparently, the absolute path of the Travis build is used (/home/travis/...) instead of the relative path to the current page.

For example, Fs Peptide (in RAM) links (in the bottom) to [this page](http://msmbuilder.org/home/travis/build/msmbuilder/msmbuilder/docs/_build/html/examples/Fs-Peptide-in-RAM/Fs-Peptide-in-RAM.ipynb

For regressionBF, I don't know how to interpret the output from model_parameters and the documentation is not giving me any hints. Is this something we should add to the docs?

BayesFactor output

# setup
set.seed(123)
library(parameters)
library(BayesFactor)
#> Loading required package: coda
#> Loading required package: Matrix
#> ************
#> Welcome to BayesFactor 0.9

Add sections on fast/slow code paths and which matrix classes work with which path, on memory use, and other internals.

We already talked about it, but it would be good to keep track with an issue.

Sphinx Gallery uses its own markup to define "input" and "output" cells. That's one more thing to "learn", and one more thing to maintain.

As an alternative, I've found nbsphinx, which allows for jupyter notebooks to be