density

D = density(x; first::Bool=true, nbins::Integer=200, bins::Vector{<:Real}=Vector{Real}(),
            bandwidth=nothing, kernel::StrSymb="normal", printbw::Bool=false,
            extend=0, plot::Bool=true, kwargs...)

Calculate the kernel density ‘d’ of a dataset X at query points determined by nbins or bins.

x: A vector, a matrix or Vector{Vector{Real}}. The density, by default, is estimated using a gaussian kernel with a width obtained with the Silverman’s rule.
nbins: points are queried between MIN(Y[:]) and MAX(Y[:]) where Y is the data vector.
bins: Calculates the density for the query points specified by BINS. The values are used as the query points directly. Default is 200 points.
bandwidth: uses the ‘bandwidth’ to calculate the kernel density. It must be a scalar. For the uniform case the bandwidth is set to 15% of the range, otherwise the bandwidth is chosen with the Silverman’s rule.
kernel: Uses the kernel function specified by KERNEL name (a string or a symbol) to calculate the density.
The kernel may be: ‘Normal’ (default) or ‘Uniform’.
printbw: Logical value indicating to print the computed value of the bandwidth.
extend: By default the density curve is computed at the bins locatins or between data extrema as mentioned above. However, this is not normally enough to go down to zero. Use this option in terms of number of bandwidth to expand de curve. e.g. extend=2
plot: Logical value indicating to plot the density curve. If false returns the GMTdataset.
kwargs: Any keyword argument accepted by plot to customize the density plot.

Returns

nothing or a GMTps and plots the density curve directly if plot=true (the fedault) and shows it is show=true.
D: A GMTdataset with two columns: first the query points and second the density values.

The version

G = density(data, x,y; weights=nothing, bandwidth=nothing, showbw=false)

Computes the smoothed kernel probability density estimate of a two-column matrix data. The estimate is based on a normal kernel function, and is evaluated at the points defined by the vectors x and y. The showbw option is used to print the bandwidth used. It returns a GMTgrid of the same size as x and y.

Example

X = [0.5*rand(20,1) 5 .+ 2.5*rand(20,1); .75 .+ 0.25*rand(10,1) 8.75 .+ 1.25*rand(10,1)];
G = GMT.density(X, -0.25:.05:1.25, 0:.1:15);
viz(G, figsize=(12,12), view=(225,30))

UndefVarError: `GMT` not defined

Stacktrace:
 [1] top-level scope
   @ In[2]:1

See other examples at

and here ?@sec-examples-of-density-plots

Source Code

This function has multiple methods:

density(D::GMTdataset, x, y; weights, bandwidth, showbw) - statplots.jl:164
density(data, x, y; weights, bandwidth, showbw, verbose) - statplots.jl:116
density(x; first, nbins, bins, bandwidth, kernel, printbw, horizontal, extend, plot, kwargs...) - statplots.jl:44

Returns

Example

Source Code

See Also