Introduce ArrayLikeVariate as common supertype for Univariate, Multiv…

…ariate and Matrixvariate (#1313)

Co-authored-by: Moritz Schauer <[email protected]>

docs/src/starting.md

@@ -49,9 +49,9 @@ julia> rand(Normal(1, 2), 100)
 
 The package contains a large number of additional distributions of three main types:
 
-* `Univariate`
-* `Multivariate`
-* `Matrixvariate`
+* `Univariate == ArrayLikeVariate{0}`
+* `Multivariate == ArrayLikeVariate{1}`
+* `Matrixvariate == ArrayLikeVariate{2}`
 
 Each type splits further into `Discrete` and `Continuous`.
 

docs/src/types.md

@@ -27,9 +27,9 @@ The `VariateForm` sub-types defined in `Distributions.jl` are:
 
 **Type** | **A single sample** | **Multiple samples**
 --- | --- |---
-`Univariate` | a scalar number | A numeric array of arbitrary shape, each element being a sample
-`Multivariate` | a numeric vector | A matrix, each column being a sample
-`Matrixvariate` | a numeric matrix | An array of matrices, each element being a sample matrix
+`Univariate == ArrayLikeVariate{0}` | a scalar number | A numeric array of arbitrary shape, each element being a sample
+`Multivariate == ArrayLikeVariate{1}` | a numeric vector | A matrix, each column being a sample
+`Matrixvariate == ArrayLikeVariate{2}` | a numeric matrix | An array of matrices, each element being a sample matrix
 
 ### ValueSupport
 

src/Distributions.jl

@@ -31,6 +31,7 @@ export
 
     # generic types
     VariateForm,
+    ArrayLikeVariate,
     ValueSupport,
     Univariate,
     Multivariate,

src/common.jl

@@ -1,13 +1,20 @@
 ## sample space/domain
 
 """
-`F <: VariateForm` specifies the form of the variate or
-dimension of a sample, univariate (scalar), multivariate (vector), matrix-variate (matrix).
+`F <: VariateForm` specifies the form or shape of the variate or a sample.
 """
 abstract type VariateForm end
-struct Univariate    <: VariateForm end
-struct Multivariate  <: VariateForm end
-struct Matrixvariate <: VariateForm end
+
+"""
+`F <: ArrayLikeVariate{N}` specifies the number of axes of a variate or
+a sample with an array-like shape, e.g. univariate (scalar, `N == 0`),
+multivariate (vector, `N == 1`) or matrix-variate (matrix, `N == 2`).
+"""
+abstract type ArrayLikeVariate{N} <: VariateForm end
+
+const Univariate    = ArrayLikeVariate{0}
+const Multivariate  = ArrayLikeVariate{1}
+const Matrixvariate = ArrayLikeVariate{2}
 
 """
 `S <: ValueSupport` specifies the support of sample elements,