* fixed multivariate

* multivariate fix for pvals > 1

* updated multivariate with hack

* bugfix

* float64 patch to multivariate's multinomial

* reinstated p.squeeze(), accidentally took it out

* updated based on failed test

* further minor fixes to multinomial

* Force push to restart Travis

* [WIP] implement random method for LKJCorr

using the algorithm in LKJ 2009(vine method based on a C-vine)

* pep8 fix

* import random method

* Revert "import random method"

This reverts commit 42e620e.

* restrict dimension n to n > 1, fix random sample for n = 2

* unify parameter naming between LKJCorr and LKJCholeskyCov

* improved random method

* Bug fix in shape

* Add test

* fix test

* pep8 clean up, fixed test warning

* improved random method

* forked R code for random method

generated Corr matrix is now positive definite.

* Use testpoint as mean in adapt_diag

* Allow missing warning in nuts check

* Fix to dtype in quadpotential

For potential usage in SMC/ABC where we need random method to generate sample. Still not encourage to use.

Extra open bracket was breaking it.

Typo in _repr_latex_ for Flat distribution

* Draft intro page

* Fixed header in intro

* Updated text with suggested edits

* Converted tabs to spaces throughout

* Reorganized sections in intro

* Minor edits to intro

* Fixed some bad rst

* Added all data to example

* Added image with example in intro

* Edited text in intro

* Fixed typo

* Added intro to index

* use Bayesian bootstrapping to compute weights

* speed up _log_post_trace

* replace bootstrapping with bootstrap

Running all cells

* fix AR(1) log likelihood

* add AR model

* fix testing floats

* resolve conflict

* Resolve conflict 2

* refine notebook

* notebook minor fix

* fix MvNormal error

(cherry picked from commit b763b08)

* add import to MvNormal directly

…hanged to 0-indexed) (pymc-devs#2493)

This is a quick patch to fix the travis fail after the change of the default init.

* Moved Constant to subclass of Continuous

* Added missing distributions from distributions api docs

* Changed Constant to be subclass of Distribution

* Reverted changes and added deprecation warning for Constant

* Moved Constant reference in init

* Moved Constant reference in init

* Added Constant to discrete.rst

* Fixed import error

* Removed duplicate/redundant warning for ConstantDist