numapprox
pade
compute a Pade approximation
Calling Sequence
Parameters
Description
Examples
pade(f, x, [m, n])
pade(f, x=a, [m, n])
pade(f, x, m)
pade(f, x=a, m)
f
-
expression representing the function to be approximated
x
the variable appearing in f
a
the finite point about which to expand in a series
m, n
desired degree of numerator and denominator, respectively
The function pade computes a Pade approximation of degree for the function with respect to the variable .
Specifically, is expanded in a Taylor (or Laurent) series about the point (if is not specified then the expansion is about the point ), to order , and then the Pade rational approximation is computed.
The Pade approximation is defined to be the rational function with and such that the Taylor (or Laurent) series expansion of has maximal initial agreement with the series expansion of . In normal cases, the series expansion agrees through the term of degree .
If the order of the lowest order term in the Laurent series is a negative integer and , then no rational approximation with a denominator of degree at most can exist, and an error is raised. If , the return value is .
If the third argument is simply an integer , then the Taylor (or Laurent) polynomial of (relative) degree is computed.
Various levels of user information will be displayed during the computation if infolevel[pade] is assigned values between and .
The command with(numapprox,pade) allows the use of the abbreviated form of this command.
Error, (in `convert/ratpoly`) no rational approximation with denominator degree <= 0
See Also
convert[ratpoly]
numapprox/hermite_pade
numapprox[chebpade]
numapprox[laurent]
taylor
Download Help Document