Visual Recurrence Analysis, version 4.0

Eugene Kononov (
Fri, 1 Oct 1999 20:06:59 -0500 (CDT)

Hello there,

This is to announce that version 4.0 of Visual Recurrence Analysis
(VRA) software for Windows 95, 98 and NT has been released.

Visual Recurrence Analysis is a software for topological analysis,
qualitative and quantitative assessment, and nonparametric prediction
of nonlinear/chaotic time series. It can detect hidden patterns and
determinism in time series using a graphical device known as the
recurrence plot. VRA first expands a given one-dimensional time
series into a higher-dimensional space, in which the dynamics of
the underlying generator takes place. VRA then constructs a
recurrence plot, which is essentially a graphical representation of
the correlation integral in such a way so that the time dependence
in the system under study is preserved.

VRA is very fast and highly interactive. Extensive online help and
references are included. A major addition to VRA v4.0 is nonlinear
time series prediction using nonparametric methods. You have a
choice of five different predictors (nearest neighbor, local constant,
kernel regression, local linear, and locally weighted linear), seven
kernel functions (Epanechnikov, Gaussian, exponential, bisquare,
tricube, inverse, triangular), five distance measures (Euclidean,
Manhattan, max norm, by cosine, by correlation), and much more.
Other new features include (modified) False Nearest Neighbors
method, recurrence histogram, and the support for the data files
in MS Excel format (in addition to plain ASCII text, comma-separated
(.csv), formatted text (.prn), and sound (.wav) formats). All charts
and graphs are now zoom-enabled and can now be saved and
printed. Recurrence quantification analysis is improved to reflect
the users' suggestions.

VRA v4.0 remains free provided that it is used for educational/
academic/research purposes. VRA v4.0 can be downloaded

If you have any comments or suggestions regarding VRA and
its use, please email me at

Thank you very much for your interest,
Eugene Kononov.