IDL Analyst Reference Guide: Random Number Generation

IMSL_FAURE_NEXT_PT

The IMSL_FAURE_NEXT_PT function computes a shuffled Faure sequence.

Syntax

Result = IMSL_FAURE_NEXT_PT(npts, state [, /DOUBLE] [, SKIP=value])

Return Value

An array of size npts by state.dim containing the npts next points in the shuffled Faure sequence.

Arguments

npts

The number of points to generate in the hyper-rectangle.

state

State structure created by a call to IMSL_FAURE_INIT.

Keywords

DOUBLE

If present and nonzero, double precision is used.

SKIP

The current point in the sequence. The sequence can be restarted by initializing a new sequence using this value for Skip, and using the same dimension for ndim.

Discussion

Discrepancy measures the deviation from uniformity of a point set.

The discrepancy of the point set:

is:

where the supremum is over all subsets of [0, 1]^d of the form:

l is the Lebesque measure, and:

is the number of the x_j contained in E.

The sequence x₁, x₂, ... of points [0,1]^d is a low-discrepancy sequence if there exists a constant c(d), depending only on d, such that:

for all n>1.

Generalized Faure sequences can be defined for any prime base b³d. The lowest bound for the discrepancy is obtained for the smallest prime b³d, so the keyword Base defaults to the smallest prime greater than or equal to the dimension. The generalized Faure sequence x₁, x₂, ..., is computed as follows:

Write the positive integer n in its b-ary expansion:

where a_i (n) are integers:

The j-th coordinate of x_n is:

The generator matrix for the series:

is defined to be:

and:

is an element of the Pascal matrix:

It is faster to compute a shuffled Faure sequence than to compute the Faure sequence itself. It can be shown that this shuffling preserves the low-discrepancy property.

The shuffling used is the b-ary Gray code. The function G(n) maps the positive integer n into the integer given by its b-ary expansion.

The sequence computed by this function is x(G(n)), where x is the generalized Faure sequence.

Example

In this example, five points in the Faure sequence are computed. The points are in the three-dimensional unit cube.

Note that IMSL_FAURE_INIT is used to create a structure that holds the state of the sequence. Each call to IMSL_FAURE_NEXT_PT returns the next point in the sequence and updates the state structure.

state = IMSL_FAURE_INIT(3)  
p = IMSL_FAURE_NEXT_PT(5, state)  
PM, p  
  
   0.333689     0.492659    0.0640654  
   0.667022     0.825992     0.397399  
   0.778133     0.270436     0.175177  
   0.111467     0.603770     0.508510  
   0.444800     0.937103     0.841843