Sampler code and example edits.

doc/examples/gk.jl

@@ -2,7 +2,8 @@ using Mamba
 
 @everywhere extensions = quote
   using Distributions
-  import Distributions: quantile
+  import Distributions: location, scale, skewness, kurtosis, minimum, maximum,
+         quantile
 
   immutable GK <: ContinuousUnivariateDistribution
     A::Float64
@@ -13,7 +14,7 @@ using Mamba
 
     function GK(A::Real, B::Real, g::Real, k::Real, c::Real)
       ## check args
-      0.0 <= c < 1.0 || throw(Argumenterror("c must be 0 <= c < 1"))
+      0.0 <= c < 1.0 || throw(ArgumentError("c must be in [0, 1)"))
 
       ## create distribution
       new(A, B, g, k, c)

doc/mcmc/distributions.rst

@@ -58,7 +58,7 @@ New known, unknown, or unnormalized univariate distributions can be created and
 
     #. Create a ``quote`` block for the new distribution.  Assign the block a variable name, say ``extensions``, preceded by the ``@everywhere`` macro to ensure compatibility when **julia** is run in multi-processor mode.
 
-    #. The *Distributions* package contains types and method definitions for new distributions.  Load the package and import the package's methods (indicated below) to be extended.
+    #. The *Distributions* package contains types and method definitions for new distributions.  Load the package and import any of its methods (indicated below) that are extended.
 
     #. Declare the new distribution subtype, say ``D``, within the block.  Any constructors explicitly defined for the subtype should accept un-typed or abstract-type (``Real``, ``AbstractArray``, or ``DenseArray``) arguments.  Implementing constructors in this way ensures that they will be callable with the *Mamba* ``Stochastic`` and ``Logical`` types.
 
@@ -147,7 +147,7 @@ New known, unknown, or unnormalized multivariate distributions can be created an
 
     #. Create a ``quote`` block for the new distribution.  Assign the block a variable name, say ``extensions``, preceded by the ``@everywhere`` macro to ensure compatibility when **julia** is run in multi-processor mode.
 
-    #. The *Distributions* package contains types and method definitions for new distributions.  Load the package and import the package's methods (indicated below) to be extended.
+    #. The *Distributions* package contains types and method definitions for new distributions.  Load the package and import any of its methods (indicated below) that are extended.
 
     #. Declare the new distribution subtype, say ``D``, within the block.  Any constructors explicitly defined for the subtype should accept un-typed or abstract-type (``Real``, ``AbstractArray``, or ``DenseArray``) arguments.  Implementing constructors in this way ensures that they will be callable with the *Mamba* ``Stochastic`` and ``Logical`` types.
 

doc/samplers/mala.jl

@@ -35,10 +35,10 @@ end
 n = 5000
 sim1 = Chains(n, 3, names = ["b0", "b1", "s2"])
 sim2 = Chains(n, 3, names = ["b0", "b1", "s2"])
-scale = 0.1
-Sigma = eye(3)
-theta1 = MALAVariate([0.0, 0.0, 0.0], scale, logfgrad)
-theta2 = MALAVariate([0.0, 0.0, 0.0], scale, Sigma, logfgrad)
+sigma = 0.1
+R = eye(3)
+theta1 = MALAVariate([0.0, 0.0, 0.0], sigma, logfgrad)
+theta2 = MALAVariate([0.0, 0.0, 0.0], sigma, R, logfgrad)
 for i in 1:n
   sample!(theta1)
   sample!(theta2)

src/samplers/amwg.jl

@@ -68,7 +68,7 @@ sample!(v::AMWGVariate; args...) = sample!(v, v.tune.logf; args...)
 function sample!(v::AMWGVariate, logf::Function; adapt::Bool=true)
   tune = v.tune
   setadapt!(v, adapt)
-  if adapt
+  if tune.adapt
     tune.m += 1
     amwg_sub!(v, logf)
     if tune.m % tune.batchsize == 0