Program pkdfnn

Version:

10.10.01

Author:

Julio José Águila Guerrero

Date:

October 20th, 2010

This parallel program implements the False Nearest Neighbor (FNN) method introduced by Kennel et al. (1991). The goal of the method is to obtain the minimal embeding dimension of a dynamical system from a time series.
In this case, the parallel implementation is based on a kd-tree data structure and the nearest neighbors searching is performed using a top-down algorithm. The kd-tree build function produces an unbalanced tree.

Note:

Last change was in October 22nd, 2010.