Stavros Macrakis wrote:
> Thanks for the bug report. Sorry Maxima gets the wrong answer here.
>
> I'm pretty sure that the problem is that Maxima simplifies (655/656)^n
> -> 655^n/656^n (which is correct), then tries to evaluate that
> expression for various float values of n. For n>109 or so, the
> intermediate values are IEEE Inf's, and so the quotient is an IEEE NaN.
> Unfortunately, Maxima doesn't know how to handle IEEE Inf/NaN, and even
> worse, does not signal an error to indicate this.
Perhaps this is a gcl problem? Both clisp and cmucl return the noun
form, with no indication of overflow or invalid operation or anything.
Ray
>
> It is also embarrassing that Maxima can't solve (655/656)^n=1/2
> symbolically.
>
> Could you please submit this bug to our bug tracking system? (Use
> bug_report()$ for instructions.)
>
> Thanks,
>
> -s
>
> On Tue, Jul 22, 2008 at 8:30 AM, Tawny Owl <tow_force at hotmail.com
> <mailto:tow_force at hotmail.com>> wrote:
>
> I am using Maxima 5.15.0 under Windows.
>
> find_root((655/656)^n = 0.5, n, 1, 1000) gives 114.8, which is
> completely wrong.
>
> Fortunately, I checked the answer (always a good habit, I know!).
>
> Either of the following expressions give the correct answer, which
> is 454.4
>
> find_root(float(655/656)^n = 0.5, n, 1, 1000);
> float(log(0.5)/log(655/656));
>
> ------------------------------------------------------------------------
> Get fish-slapping on Messenger! Play Now
> <http://clk.atdmt.com/UKM/go/101719805/direct/01/>;
>
> _______________________________________________
> Maxima mailing list
> Maxima at math.utexas.edu <mailto:Maxima at math.utexas.edu>
> http://www.math.utexas.edu/mailman/listinfo/maxima
>
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Maxima mailing list
> Maxima at math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima