Package csb :: Package statistics :: Module rand

[frames] | no frames]

Module rand

source code

Random number generators

Functions

gen_inv_gaussian(a, b, p, burnin=10)
Sampler based on Gibbs sampling.

source code

inv_gaussian(mu=1.0, _lambda=1.0, shape=None)
Generate random samples from inverse gaussian.

source code

probability_transform(shape, inv_cum, cum_min=0.0, cum_max=1.0)
Generic sampler based on the probability transform.

source code

3 x 3 numpy array

random_rotation(A, n_iter=10, initial_values=None)
Generation of three-dimensional random rotations in fitting and matching problems, Habeck 2009.

source code

sample_dirichlet(alpha, n_samples=1)
Sample points from a dirichlet distribution with parameter alpha.

source code

sample_from_histogram(p, n_samples=1)
returns the indice of bin according to the histogram p

source code

numpy array

sample_sphere3d(radius=1.0, n_samples=1)
Sample points from 3D sphere.

source code

truncated_gamma(shape=None, alpha=1.0, beta=1.0, x_min=None, x_max=None)
Generate random variates from a lower-and upper-bounded gamma distribution.

source code

truncated_normal(shape=None, mu=0.0, sigma=1.0, x_min=None, x_max=None)
Generates random variates from a lower-and upper-bounded normal distribution

source code

Variables
	__package__ = `None` hash(x)

Function Details

gen_inv_gaussian(a, b, p, burnin=10)

source code

Sampler based on Gibbs sampling. Assumes scalar p.

probability_transform(shape, inv_cum, cum_min=0.0, cum_max=1.0)

source code

Generic sampler based on the probability transform.

Parameters:

shape - shape of the random sample
inv_cum - inversion of the cumulative density function from which one seeks to sample
cum_min - lower value of the cumulative distribution
cum_max - upper value of the cumulative distribution

Returns:

random variates of the PDF implied by the inverse cumulative distribution

random_rotation(A, n_iter=10, initial_values=None)

source code

Generation of three-dimensional random rotations in fitting and matching problems, Habeck 2009.

Generate random rotation R from:

   exp(trace(dot(transpose(A), R)))

Parameters:

A (3 x 3 numpy array) - generating parameter
n_iter (integer) - number of gibbs sampling steps
initial_values (tuple) - initial euler angles alpha, beta and gamma

Returns: 3 x 3 numpy array

sample_dirichlet(alpha, n_samples=1)

source code

Sample points from a dirichlet distribution with parameter alpha.

Parameters:

alpha (array) - alpha parameter of a dirichlet distribution

sample_from_histogram(p, n_samples=1)

source code

returns the indice of bin according to the histogram p

Parameters:

p (numpy.array) - histogram
n_samples (integer) - number of samples to generate

sample_sphere3d(radius=1.0, n_samples=1)

source code

Sample points from 3D sphere.

Parameters:

radius (float) - radius of the sphere
n_samples (int) - number of samples to return

Returns: numpy array

n_samples times random cartesian coordinates inside the sphere

truncated_gamma(shape=None, alpha=1.0, beta=1.0, x_min=None, x_max=None)

source code

Generate random variates from a lower-and upper-bounded gamma distribution.

Parameters:

shape - shape of the random sample
alpha - shape parameter (alpha > 0.)
beta - scale parameter (beta >= 0.)
x_min - lower bound of variate
x_max - upper bound of variate

Returns:

random variates of lower-bounded gamma distribution

truncated_normal(shape=None, mu=0.0, sigma=1.0, x_min=None, x_max=None)

source code

Generates random variates from a lower-and upper-bounded normal distribution

Parameters:

shape - shape of the random sample
mu - location parameter
sigma - width of the distribution (sigma >= 0.)
x_min - lower bound of variate
x_max - upper bound of variate

Returns:

random variates of lower-bounded normal distribution