David Mayer-Foulkes’ Download Page The program U-Stat can be downloaded here The PAHO report on investment in health can be downloaded here
INVESTMENT IN HEALTH AND ECONOMIC GROWTH GENERALIZED FAST ALGORITHM FOR BDS-TYPE
STATISTICS U-STAT - DESCRIPTION OF THE PROGRAM
The research
project "Health, Growth and Income Distribution in Latin America and the
Caribbean: A Study of Determinants and regional and Local Behavior " was
the winner of the 1997 Regional Research Competition on Investment in Health
and Economic Growth of the Pan American Health Organization. To download the
report in English or in Spanish, written by Mayer, D.; Mora, H. Cermeño, R.
and Duryea, S, use one of the links below.
To download
the article Mayer, D., "A generalized fast algorithm for BDS-type
Statistics", The program
U-Stat can be downloaded here
U-Stat is a
Windows program to calculate several statistics based on the
multi-dimensional distance histogram - The Correlation Dimension (Grassberger and Procaccia, 1983)
- The BDS statistic (Brock, 1986a, 1986b; Brock, Dechert, Sheinkman
and LeBaron, 1996)
- The Simple Non-parametric Test (Mizrach, 1991, 1992, 1994)
- The Statistical Correlation Dimension or Correlation Dimension
Ratio (Mayer, 1995; Mayer and Feliz 1996)
- Homogenized Integral U-Statistics (Mayer, 1998)
- Further applications to vector series (Mayer, to be reported)
All
confidence intervals are determined by using Brock’s (1986a, 1986b)
reshuffling test. We classify
the statistics into orders 1 and 2, according to whether - The generalized fast algorithm (Mayer, 2000)
- An algorithm generalizing LeBaron’s (1997) fast algorithm using
sorting which is of order
*N*when the calculation is restricted to small values of*e*(Mayer, to be reported) - An algorithm implementing Grassberger’s (1990) box-assisted algorithm
(including the calculation of
*C*(1,*e*) and the use of tao) which is also of order*N*when the calculation is restricted to small values of*e*(Mayer, to be reported)
The author
thanks CONACYT for the funding received for the development of this program
through the research project "Estadísticos U para Pruebas de
Independencia No Lineales" (project number 3416P-S9607). Several
additions to U-Stat are presently being written. Thus if you mail me at david.mayer@cide.edu. I will notify you as soon as any changes are
implemented. The program
may be freely distributed, so long as its use is cited and acknowledged in
full and the terms of the license detailed in the program are kept (all
rights reserved). Please report
on any problems, so that the program may be improved. At this stage the
program has not been run on many machines so some problems may crop up. Any
suggestions and bug reports are welcome. Once you
download the file ustat.zip, decompress it in a directory c:\ustat and follow
these instructions. 1) Copy the files VCF132.OCX, OC30.DLL,
MFCANS32.DLL, MSVCRT20.DLL to the directory c:/Windows/System. If your system
has more recent versions of some of these files, do not substitute them! 2) Using the option Run in the Start menu,
execute the command Regsvr32.exe Vcf132.ocx. If this command returns a
success message, the program will be ready to run by double clicking on
ustat.exe. Brock, W. A.
(1986a), "Distinguishing Random and Deterministic Systems: Abridged
Version", Brock, W. A.
(1986b), "Theorems on Distinguishing Deterministic from Random
Systems", Brock, W. A.;
Dechert, W. D.; Sheinkman, J.A. and LeBaron, B. (1996), "A Test for
Independence Based on the Correlation Dimension", Dahlquist,
G., and Björck, A. (1974), Grassberger,
Peter and Itamar Procaccia (1983), "Measuring the strangeness of strange
attractors", Grassberger,
Peter (1990), "An optimized box-assisted algorithm for fractal
dimensions'', LeBaron,
Blake (1997) "A Fast Algorithm for the BDS Statistic'', Mayer-Foulkes,
D. (2000), "A generalized fast algorithm for BDS-type Statistics",
forthcoming in Mayer-Foulkes,
D. (1998), "Homogenized Integral U-Statistics for Test of
Non-Linearity", Documento de Trabajo del CIDE, División de Economía, No.
118. Mayer-Foulkes,
D. (1995), "A statistical correlation dimension", Mayer-Foulkes,
D., and Raúl Anibal Feliz (1996), "Nonlinear dynamics in the stock
exchange", Mizrach, B.
(1991). "A simple Nonparametric test for independence." Working
paper, Department of Finance, the Wharton School. Mizrach, B.
(1992). "Multivariate nearest-neighbour Forecasts of EMS Exchange
Rates." Mizrach, B.
(1994). "Using U-statistics to detect business cycle
non-linearities." Chapter 14 in Willi Semmler (ed) |