R help - Sep 2010 - Inaccuracy of kummerU (fAsianOptions) (Tricomi function)

Hello,
I need to use the confluent function of second kind, also known as 
Tricomi function. It is implemented as kummerU() function in 
fAsianOptions package, but I've found very inaccurate values, comparing 
with those provided by Mathematica. I think Mathematica values are OK 
because kummerU values leads to negative probabilities.
For example, if you try the kummerU() function example, you obtain:
x = c(0.001, 0.01, 0.1, 1, 10, 100, 1000)
         a = 1/3; b = 2/3
         U = Re ( kummerU(x, a = a, b = b) )
         cbind(x, U)
          x             U  Mathematica
[1,] 1e-03  1.827638e+00  1.82764
[2,] 1e-02  1.659957e+00  1.65996
[3,] 1e-01  1.341021e+00  1.34102
[4,] 1e+00  8.745968e-01  0.874597
[5,] 1e+01  4.548890e-01  0.454805
[6,] 1e+02  1.731286e+35  0.214970
[7,] 1e+03 -1.318568e+76  0.0999778

I've added a third column in the table with Mathematica values, where 
you can see the great differences in the last values. I have a lot of 
examples of these differences, with non very strange parameters values.
I've also tried to define the Tricomi function in relation to the usual 
confluent function given by genhypergeo(), but errors are even greater.
Does anybody know another function or another way to define an accurate 
version of Tricomi function?

-- 
Dr. Antonio José Sáez Castillo
Dpto. de Estadística e Investigación Operativa
Escuela Politécnica Superior de Linares
Universidad de Jaén
C/ Alfonso X El Sabio 28, 23700 Linares (Jaén) ESPAÑA
Tlf. y FAX +34 953 648578


	[[alternative HTML version deleted]]