Update kernelpdmat to accept vector of vectors (#252)

Author: theogf

Project.toml

@@ -1,6 +1,6 @@
 name = "KernelFunctions"
 uuid = "ec8451be-7e33-11e9-00cf-bbf324bd1392"
-version = "0.8.23"
+version = "0.8.24"
 
 [deps]
 Compat = "34da2185-b29b-5c13-b0c7-acf172513d20"

src/matrix/kernelpdmat.jl

@@ -3,13 +3,20 @@ using .PDMats: PDMat
 export kernelpdmat
 
 """
-    Compute a positive-definite matrix in the form of a `PDMat` matrix see [PDMats.jl](https://github.com/JuliaStats/PDMats.jl)
-    with the cholesky decomposition precomputed.
-    The algorithm recursively tries to add recursively a diagonal nugget until positive
-    definiteness is achieved or that the noise is too big.
+    kernelpdmat(k::Kernel, X::AbstractMatrix; obsdim::Int=2)
+    kernelpdmat(k::Kernel, X::AbstractVector)
+
+Compute a positive-definite matrix in the form of a `PDMat` matrix see [PDMats.jl](https://github.com/JuliaStats/PDMats.jl)
+with the cholesky decomposition precomputed.
+The algorithm recursively tries to add recursively a diagonal nugget until positive
+definiteness is achieved or until the noise is too big.
 """
 function kernelpdmat(κ::Kernel, X::AbstractMatrix; obsdim::Int=defaultobs)
-    K = kernelmatrix(κ, X; obsdim=obsdim)
+    return kernelpdmat(κ, vec_of_vecs(X; obsdim=obsdim))
+end
+
+function kernelpdmat(κ::Kernel, X::AbstractVector)
+    K = kernelmatrix(κ, X)
     Kmax = maximum(K)
     α = eps(eltype(K))
     while !isposdef(K + α * I) && α < 0.01 * Kmax
@@ -24,7 +31,3 @@ function kernelpdmat(κ::Kernel, X::AbstractMatrix; obsdim::Int=defaultobs)
     end
     return PDMat(K + α * I)
 end
-
-function kernelpdmat(κ::Kernel, X::AbstractVector{<:Real}; obsdim=defaultobs)
-    return kernelpdmat(κ, reshape(X, 1, :); obsdim=2)
-end

test/matrix/kernelpdmat.jl

@@ -1,12 +1,15 @@
 @testset "kernelpdmat" begin
     rng = MersenneTwister(123456)
     A = rand(rng, 10, 5)
+    vecA = (RowVecs(A), ColVecs(A))
+    a = rand(rng, 10)
     k = SqExponentialKernel()
     for obsdim in [1, 2]
         @test all(
             Matrix(kernelpdmat(k, A; obsdim=obsdim)) .≈
             Matrix(PDMat(kernelmatrix(k, A; obsdim=obsdim))),
         )
+        @test kernelpdmat(k, vecA[obsdim]) == kernelpdmat(k, A; obsdim=obsdim)
         # @test_throws ErrorException kernelpdmat(k,ones(100,100),obsdim=obsdim)
     end
 end