Merge branch 'master' into dw/abstractffts

Project.toml

@@ -14,8 +14,10 @@ StatsBase = "2913bbd2-ae8a-5f71-8c99-4fb6c76f3a91"
 AbstractFFTs = "0.5, 1"
 Distributions = "0.23, 0.24, 0.25"
 DocStringExtensions = "0.8, 0.9"
-Interpolations = "0.9, 0.10, 0.11, 0.12, 0.13, 0.14"
+FFTW = "1"
+Interpolations = "0.9, 0.10, 0.11, 0.12, 0.13, 0.14, 0.15"
 StatsBase = "0.33, 0.34"
+Test = "<0.0.1, 1"
 julia = "1"
 
 [extras]

src/univariate.jl

@@ -147,6 +147,20 @@ function conv(k::UnivariateKDE, dist::UnivariateDistribution)
 end
 
 # main kde interface methods
+
+"""
+
+	kde(data; kwargs...)
+	kde((xdata, ydata); kwargs...)
+	
+Kernel density estimation method. Returns 1D or 2D KDE object. The grid used and the values of the estimated density can be obtained from fields `.x` and `.density` respectively. To obtain kde values at points different than the initial grid use the `pdf` method.
+
+The keyword arguments are
+* `boundary`: the lower and upper limits of the kde, tuple in 1D case, tuple of tuples in 2D case, 
+* `npoints`: the number of interpolation points to use,
+* `kernel = Normal`: the distributional family from [Distributions.jl](https://github.com/JuliaStats/Distributions.jl),
+* `bandwidth`: the bandwidth of the kernel; default is calculated using Silverman's rule. 
+"""
 function kde(data::AbstractVector{<:Real}, weights::Weights, midpoints::R, dist::UnivariateDistribution) where R<:AbstractRange
     k = tabulate(data, midpoints, weights)
     conv(k,dist)