test10.cpp File Reference

---

Detailed Description

Test of KernelDensityPdf and KDTree.

This test:

• creates a random multivariate Gaussian mixture,
• samples from this mixture and constructs a ,
• creates a kernel density estimate from this tree and performs various density calculations using this.

Results are as follows:

Mixture mean
[2](-0.0774694,-0.318281)
Mixture covariance
[2,2]((0.946293,-0.175484),(-0.175484,2.43054))
Sample mean
[2](-0.0770399,-0.340447)
Sample covariance
[2,2]((0.927742,-0.172208),(-0.172208,2.54067))
Kernel density mean
[2](-0.0770399,-0.340447)
Kernel density tree covariance
[2,2]((0.927742,-0.172208),(-0.172208,2.54067))
Kernel density sample mean
[2](-0.118136,-0.25066)
Kernel density sample covariance
[2,2]((0.888061,-0.215785),(-0.215785,2.454))
Kernel density importance sample mean
[2](-0.0458998,-0.358719)
Kernel density importance sample covariance
[2,2]((1.19893,-0.684056),(-0.684056,3.15693))

Original Gaussian mixture

Kernel density approximation

Definition in file test10.cpp.

Go to the source code of this file.

Functions
aux::GaussianPdf	createRandomGaussian (const unsigned int M, const double minMean=-1.0, const double maxMean=1.0, const double minCov=-1.0, const double maxCov=1.0)
	Create random Gaussian distribution.
int	main (int argc, char *argv[])
	Run tests.
Variables
unsigned int	M = 2
	Dimensionality of the distribution.
unsigned int	COMPONENTS = 4
	Number of components in the Gaussian mixture.
unsigned int	P = 1000
	Number of samples to take.
unsigned int	RES = 200
	Resolution of plots.
double	H = 0.25 * std::pow((double)4/(P*(M+2)), (double)1/(M+4))
	Scaling parameter.

---

Function Documentation

aux::GaussianPdf createRandomGaussian	(	const unsigned int	M,
		const double	minMean = -1.0,
		const double	maxMean = 1.0,
		const double	minCov = -1.0,
		const double	maxCov = 1.0
	)

Create random Gaussian distribution.

Parameters:

	M	Dimensionality of the Gaussian.
	minMean	Minimum value of any component of the mean.
	maxMean	Maximum value of any component of the mean.
	minCov	Minimum value of any component of the covariance.
	maxCov	Maximum value of any component of the covariance.

Returns:

Gaussian with given dimensionality, with mean and covariance randomly generated uniformly from within the given bounds.

Definition at line 83 of file test10.cpp.

int main	(	int	argc,
		char *	argv[]
	)

Run tests.

Definition at line 112 of file test10.cpp.

---

Variable Documentation

unsigned int COMPONENTS = 4

Number of components in the Gaussian mixture.

Definition at line 53 of file test10.cpp.

double H = 0.25 * std::pow((double)4/(P*(M+2)), (double)1/(M+4))

Scaling parameter.

Definition at line 68 of file test10.cpp.

unsigned int M = 2

Dimensionality of the distribution.

Definition at line 48 of file test10.cpp.

unsigned int P = 1000

Number of samples to take.

Definition at line 58 of file test10.cpp.

unsigned int RES = 200

Resolution of plots.

Definition at line 63 of file test10.cpp.

---