chrono::ChContinuumDistribution Class Reference

Description

Class that can be used to generate sample numbers according to a probability distribution.

Probability distribution is defined with x,y points, at least a dozen of pairs to be more precise in the reconstruction of probability. It uses the "Smirnov transform" (inverse probability integral transform)

#include <ChRandom.h>

Inheritance diagram for chrono::ChContinuumDistribution:
Collaboration diagram for chrono::ChContinuumDistribution:

Public Member Functions

 ChContinuumDistribution (ChVectorDynamic<> &x, ChVectorDynamic<> &y)
 Create an object that can be used to generate sample numbers according to a generic probability distribution. More...
 
virtual double GetRandom () override
 Compute a random value whose probability is the probability curve that has been entered with x,y points during the creation of this object.
 
const ChVectorDynamic & GetProbabilityXPoints () const
 
const ChVectorDynamic & GetProbabilityYPoints () const
 
const ChVectorDynamic & GetProbabilityCDFCumulativeX () const
 
const ChVectorDynamic & GetProbabilityCDFCumulativeY () const
 

Constructor & Destructor Documentation

◆ ChContinuumDistribution()

chrono::ChContinuumDistribution::ChContinuumDistribution ( ChVectorDynamic<> &  x,
ChVectorDynamic<> &  y 
)

Create an object that can be used to generate sample numbers according to a generic probability distribution.

The probability distribution is a curve represented by simplified x,y pairs of points. The integral of the probability curve must be unit, i.e normalized (but if not, a normalization will be enforced) Note: too few points means approximate results, but too many points might give a small performance overhead when calling GetRandom().


The documentation for this class was generated from the following files:
  • /builds/uwsbel/chrono/src/chrono/core/ChRandom.h
  • /builds/uwsbel/chrono/src/chrono/core/ChRandom.cpp