MathGroup Archive 2009

[Date Index] [Thread Index] [Author Index]

Search the Archive

Precision of AiryAi[0.0]

  • To: mathgroup at smc.vnet.net
  • Subject: [mg103466] Precision of AiryAi[0.0]
  • From: Mark McClure <mcmcclur at unca.edu>
  • Date: Wed, 23 Sep 2009 07:25:11 -0400 (EDT)

Running V7.0.1 on my 64 bit MacBook Pro.

The result of the simple computation AiryAi[0.0] returns a
result with a precision of 15.9, rather than the expected
MachinePrecision.  That is, we get:

Precision[AiryAi[0.0]]
15.9546

Rather than MachinePrecision as in

Precision[BesselJ[2, 0.0]]
MachinePrecision


Does anyone have an explanation for this?

This occurs with all of the Airy related functions, but
none of the other special functions I've tried have this
strange behavior.  Of course, some computations (like
2^2000.0) can lead to arbitrary precision results, simply
because the result can't be expressed in IEEE floating
point arithmetic.  AiryAi[0.0] is about 0.355, however,
so it seems quite strange.

Incidentally, the problem arose as I was studying the
complex dynamics of Airy functions.  My programs were
running slower than I expected.  I traced the problem
down to this unexpected appearance arbitrary precision
numbers.

Mark McClure


  • Prev by Date: Re: Using Mathematica 5.2 and 7.01 with separate
  • Next by Date: Huge memory leaks when importing two-dimensional array in Mathematica
  • Previous by thread: Re: Re: Using Mathematica 5.2 and 7.01 with separate
  • Next by thread: Re: Precision of AiryAi[0.0]