5
$\begingroup$

Bug introduced in 9.0.0 and fixed in 10.0.1


Limit[(Log[(3+Sqrt[5])/2]/(2*Log[(1+Sqrt[5])/2]))^(-1-2*n), n -> Infinity]

Mathematica (wrong) output:

0

But this is correct:

FullSimplify[(Log[(3+Sqrt[5])/2]/(2*Log[(1+Sqrt[5])/2]))^(-1-2*n)]
1

And also this is correct:

Limit[(Log[4]/(2*Log[2]))^(-1-2*n), n -> Infinity]
1
$\endgroup$
  • 4
    $\begingroup$ Version 10.1 under Windows x64 gives 1 for the first input line. This is probably a bug that has been corrected. $\endgroup$ – Mr.Wizard May 22 '15 at 10:10
  • $\begingroup$ Thank you, I tested it under version 10.0.0.0 $\endgroup$ – Vaclav Kotesovec May 22 '15 at 10:17
  • 1
    $\begingroup$ Gives 0 in version 9.0.1, and returns unevaluated in previous versions. Looks like the bug was introduced in version 9. $\endgroup$ – Oleksandr R. May 22 '15 at 11:22
  • $\begingroup$ Version 10.0.2 also gives 1. Can anyone test 10.0.1? $\endgroup$ – Mr.Wizard May 22 '15 at 11:41
  • 1
    $\begingroup$ 10.0.1 gives 1. $\endgroup$ – N.J.Evans May 22 '15 at 13:52
1
$\begingroup$

Summarizing the comments, this bug has been fixed as of version 10.0.1.

Limit[(Log[(3+Sqrt[5])/2]/(2*Log[(1+Sqrt[5])/2]))^(-1-2*n), n -> Infinity]    

(* 1 *)
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.