login  home  contents  what's new  discussion  bug reports help  links  subscribe  changes  refresh  edit

 Submitted by : (unknown) at: 2007-11-17T22:13:08-08:00 (10 years ago) Name : Axiom Version : default friCAS-20090114 Axiom-20050901 OpenAxiom-20091012 OpenAxiom-20110220 OpenAxiom-Release-141 Category : Axiom Aldor Interface Axiom Compiler Axiom Library Axiom Interpreter Axiom Documentation Axiom User Interface building Axiom from source lisp system MathAction Doyen CD Reduce Axiom on Windows Axiom on Linux Severity : critical serious normal minor wishlist Status : open closed rejected not reproducible fix proposed fixed somewhere duplicate need more info Optional subject :   Optional comment :

Another problem I have is that taking the limit of an expression containing (-1)^n always returns "failed", where my TI-89 Titanium calculator will give a finite limit. For example:

limit( 2 + (-2/%pi)^n, n=%plusInfinity ) ===> "failed"

... but the TI-89t returns 2.

This version of Axiom (FriCAS) gives:

fricas
)version
Value = "FriCAS 1.2.1 compiled at Monday July 1, 2013 at 23:26:54 "
limit( 2 + (-2/%pi)^n, n=%plusInfinity )
 (1)
Type: Union("failed",...)

The TI-89t says that the limit of (-1)^n as n approaches infinity is -1, implying that it believes that infinity is an odd number. That kind of makes sense to me, since if you divide infinity in half, you still have infinity, and you keep adding 1 to get to infinity, making it odd. If infinity is even then the answer should be 1, and if we can't know if infinity is even or odd, then the answer is uncertain or undefined.

On the other hand, the TI-89t says that lim ( (-1)^n * (n + 1)/n ) is undefined. But it already told me that lim (-1)^n = -1, and that lim (n + 1)/n = 1. If the limit of a product is the product of the limits of the factors, then lim ( (-1)^n * (n + 1)/n ) should be -1, right?

So, who's right?

An answer to my question, from William Sit. --unknown, Sun, 13 Nov 2005 11:27:38 -0600 reply
William Sit responded via emai, and wrote, in part:

Mathematically, the limit is 2 since (-2/%pi) has absolute value less than 1, and hence (-2/%pi)^n converges to 0. So TI-89t is correct and Axiom is wrong.

In the second case, TI-89t is wrong to say that limit (-1)^n is -1. The limit does not exist because the sequence (-1)^n oscillates between 1 and -1. There is no number L (the assumed limit) such that given any epsilon > 0, there is a natural number N such that |(-1)^n - L| < epsilon for all n > N.

At least Axiom gets this right:

fricas
limit( (-1)^n,  n=%plusInfinity )
 (2)
Type: Union("failed",...)
fricas
limit( (n+ 1)/n,  n=%plusInfinity )
 (3)
Type: Union(OrderedCompletion?(Fraction(Polynomial(Integer))),...)
fricas
limit( (-1)^n * (n + 1)/n,  n=%plusInfinity )
 (4)
Type: Union("failed",...)

Category: Axiom Mathematics => Axiom Library

Name: #234 limit((-1/2)^n,n=%plusInfinity) => #234 limit(-1/2)^n

 Subject:   Be Bold !! ( 13 subscribers )