kernelDensity

Purpose

To compute kernel density estimate and plot.

Format

out = kernelDensity(dataset[, kernel, bw, c0])

out = kernelDensity(dataset, formula[, kernel, bw, c0])

Parameters

• dataset (String) – Name of datafile, dataframe, or data matrix.

• formula – Optional argument, formula string for loading variables from datafile. Default is to load all variables.

• kernel (Scalar) –

  Optional argument, type of kernel. Default = 1.

  1	Normal.
  2	Epanechnikov.
  3	Biweight.
  4	Triangular
  5	Rectangular.
  6	Truncated normal.
  7	Parzen.
  8	Cosine.
  9	Triweight.
  10	Tricube.
  11	Logistic.
  12	Sigmoid.
  13	Silverman.

  If kernel is scalar, the kernel is the same for all parameters. Default = Normal density.

• bw (Scalar, Vector, or Matrix) – Optional argument, smoothing coefficient (bandwidth). Should be ExE compatible with the dataset. If bw is scalar, the smoothing coefficient will be the same for every data observation. If bw is zero, optimal smoothing coefficient will be computed. If bw` is a matrix, the smoothing coefficient will be different for each observation. Default = 0.

• c0 (Structure) –

  Instance of kernelDensityControl structure containing the following members:

  c0.varNames	String array, variable name(s). Default is dataframe headers if included, "X1", "X2", … otherwise.
  c0.NumPoints	Scalar, number of points to be computed for plots.
  c0.EndPoints	Kx2 matrix, lower (in first column) and upper (in second column) endpoints of density. Default is minimum and maximum, respectively, of the parameter values. If 1x2 matrix, endpoints will be the same for all parameters.
  c0.Truncate	Kx2 matrix, lower (in first column) and upper (in second column) truncation limits for truncated normal kernel. If 1x2 matrix, truncations limits will be the same for all plots. Default is minimum and maximum, respectively.
  c0.plotFunctionPtr	Scalar, a pointer to a user-defined function used to specify format modifications to the default plot format.

Returns

out (Structure) –

Instance of KernelDensityResults structure

out.px	Matrix, abscissae.
out.py	Matrix, ordinates.
out.sm	Kx1, or Nxk, or Nx1 smoothing coefficients.

Examples

Basic usage with dataframe

/*
** Example One:
** Basic usage
*/

// Load data
fname = getGAUSSHome $+ "examples/winevolatileacidity.csv";
data = loadd(fname);

// Call kernelDensity function
// with default normal density
struct kernelDensityResults krslt1;
krslt1 = kernelDensity(data[., "volatile acidity"]);

Basic usage with filename

/*
** Example Two:
** Basic usage with filename
*/

// Load data
fname = getGAUSSHome $+ "examples/winevolatileacidity.csv";

// Call kernelDensity function
// with default normal density
call kernelDensity(fname, "volatile acidity");

Basic usage with multiple kernels

/*
** Example Three:
** Multiple kernels
** on same plot
*/

struct kernelDensityResults krslt2;
krslt2 = kernelDensity(data[., "volatile acidity"], 1|2|3);

Basic usage with multiple series

/*
** Example Four:
** Multiple series on different plots
*/

struct kernelDensityResults krslt3;
krslt3 = kernelDensity(data, 1);

Modifying plot with plotControl pointer

/*
** Example Five:
** Modifying plot
*/

struct kernelDensityControl kctl;
kctl = kernelDensityControlCreate();

kctl.plotFunctionPtr = &myPlotCustomizations;

// Basic example of cmlmtKerneldensity plot customization
proc (1) = myPlotCustomizations(struct plotControl myPlot);

  //Make all lines 2px thick
  plotSetGridPen(&myPlot, "major", 1, "dimgray", 2);

  //Return modified plotControl structure
  retp(myPlot);

endp;

// Set kernel to normal
kernel = 1;

// Automatically compute bandwidth
bw = 0;

// Call kernelDensity with
// control structure
struct kernelDensityResults krslt4;
krslt4 = kernelDensity(data[., "volatile acidity"], kernel, bw, kctl);

previous

keep (dataloop)

next

keyav