Merge pull request #37 from JuliaStats/sb/interpolations

move from Grid.jl to Interpolations.jl

Author: simonbyrne

.travis.yml

@@ -3,7 +3,6 @@ os:
   - linux
   - osx
 julia:
-  - 0.3
   - 0.4
   - 0.5
   - nightly

README.md

@@ -2,7 +2,6 @@
 
 [![Build Status](https://travis-ci.org/JuliaStats/KernelDensity.jl.svg?branch=master)](https://travis-ci.org/JuliaStats/KernelDensity.jl)
 [![Coverage Status](https://coveralls.io/repos/github/JuliaStats/KernelDensity.jl/badge.svg)](https://coveralls.io/github/JuliaStats/KernelDensity.jl)
-[![KernelDensity](http://pkg.julialang.org/badges/KernelDensity_0.3.svg)](http://pkg.julialang.org/?pkg=KernelDensity&ver=0.3)
 [![KernelDensity](http://pkg.julialang.org/badges/KernelDensity_0.4.svg)](http://pkg.julialang.org/?pkg=KernelDensity&ver=0.4)
 [![KernelDensity](http://pkg.julialang.org/badges/KernelDensity_0.5.svg)](http://pkg.julialang.org/?pkg=KernelDensity)
 
@@ -78,7 +77,7 @@ e.g. `boundary` now takes a tuple of tuples `((xlo,xhi),(ylo,yhi))`.
 The KDE objects are stored as gridded density values, with attached
 coordinates. These are typically sufficient for plotting (see below), but
 intermediate values can be interpolated using the
-[Grid.jl](https://github.com/timholy/Grid.jl) package via the `pdf` method
+[Interpolations.jl](https://github.com/tlycken/Interpolations.jl) package via the `pdf` method
 (extended from Distributions.jl).
 
 ```
@@ -89,15 +88,12 @@ pdf(k::BivariateKDE, x, y)
 where `x` and `y` are real numbers or arrays.
 
 If you are making multiple calls to `pdf`, it will be more efficient to
-construct an intermediate `InterpKDE` to store the intermediate `InterpGrid`
-object:
+construct an intermediate `InterpKDE` to store the interpolation structure:
 
 ```
 ik = InterpKDE(k)
 pdf(ik, x)
 ```
 
-`InterpKDE` can also take additional arguments specifying the
-`BoundaryCondition` (default=`BCnan`) and `InterpType` (default=
-`InterpQuadratic`).
+`InterpKDE` will pass any extra arguments to `interpolate`.
 

REQUIRE

@@ -1,6 +1,6 @@
-julia 0.3
+julia 0.4
 StatsBase
 Distributions
 Optim
-Grid
+Interpolations
 Compat

src/KernelDensity.jl

@@ -5,7 +5,7 @@ module KernelDensity
 using StatsBase
 using Distributions
 using Optim
-using Grid
+using Interpolations
 using Compat
 
 import Base: conv
@@ -14,6 +14,8 @@ import Distributions: twoπ, pdf
 
 export kde, kde_lscv, UnivariateKDE, BivariateKDE, InterpKDE, pdf
 
+abstract AbstractKDE
+
 include("univariate.jl")
 include("bivariate.jl")
 include("interp.jl")

src/bivariate.jl

@@ -1,5 +1,5 @@
 # Store both grid and density for KDE over R2
-type BivariateKDE{Rx<:Range,Ry<:Range}
+type BivariateKDE{Rx<:Range,Ry<:Range} <: AbstractKDE
     x::Rx
     y::Ry
     density::Matrix{Float64}

src/interp.jl

@@ -1,19 +1,30 @@
-type InterpKDE{K,G}
+import Interpolations: interpolate
+
+type InterpKDE{K,I} <: AbstractKDE
     kde::K
-    grid::G
+    itp::I
+    InterpKDE(kde,itp) = new(kde,itp)
 end
 
-function InterpKDE{BC<:BoundaryCondition, IT<:InterpType}(k::UnivariateKDE, bc::Type{BC}=BCnan,it::Type{IT}=InterpQuadratic)
-    g = CoordInterpGrid(k.x, k.density, bc, it)
-    InterpKDE(k,g)
-end
-function InterpKDE{BC<:BoundaryCondition, IT<:InterpType}(k::BivariateKDE, bc::Type{BC}=BCnan,it::Type{IT}=InterpQuadratic)
-    g = CoordInterpGrid((k.x,k.y),k.density,bc,it)
-    InterpKDE{typeof(k),typeof(g)}(k,g)
+
+function InterpKDE(kde::UnivariateKDE, opts...)
+    itp_u = interpolate(kde.density, opts...)
+    itp = scale(itp_u, kde.x)
+    InterpKDE{typeof(kde),typeof(itp)}(kde, itp)
 end
+InterpKDE(kde::UnivariateKDE) = InterpKDE(kde, BSpline(Quadratic(Line())), OnGrid())
 
 
-pdf(ik::InterpKDE,x...) = ik.grid[x...]
+function InterpKDE(kde::BivariateKDE, opts...)
+    itp_u = interpolate(kde.density,opts...)
+    itp = scale(itp_u, kde.x, kde.y)
+    InterpKDE{typeof(kde),typeof(itp)}(kde, itp)
+end
+InterpKDE(kde::BivariateKDE) = InterpKDE(kde::BivariateKDE, BSpline(Quadratic(Line())), OnGrid())
+
+pdf(ik::InterpKDE,x::Real...) = ik.itp[x...]
+pdf(ik::InterpKDE,xs::AbstractVector) = [ik.itp[x] for x in xs]
+pdf(ik::InterpKDE,xs::AbstractVector,ys::AbstractVector) = [ik.itp[x,y] for x in xs, y in ys]
 
 pdf(k::UnivariateKDE,x) = pdf(InterpKDE(k),x)
 pdf(k::BivariateKDE,x,y) = pdf(InterpKDE(k),x,y)

src/univariate.jl

@@ -1,5 +1,5 @@
 # Store both grid and density for KDE over the real line
-type UnivariateKDE{R<:Range}
+type UnivariateKDE{R<:Range} <: AbstractKDE
     x::R
     density::Vector{Float64}
 end

test/interp.jl

@@ -1,6 +1,5 @@
 using Base.Test
 using KernelDensity
-using Grid
 
 X = randn(100)
 Y = randn(100)

test/runtests.jl

@@ -1,6 +1,7 @@
 tests = [
          "univariate",
          "bivariate",
+         "interp",
          ]
 
 println("Running tests:")