Services & Resources / Wolfram Forums / MathGroup Archive
-----

MathGroup Archive 2009

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

Search the Archive

Re: NMinimize Bug in Mathematica 7.0?

  • To: mathgroup at smc.vnet.net
  • Subject: [mg96270] Re: NMinimize Bug in Mathematica 7.0?
  • From: dh <dh at metrohm.com>
  • Date: Tue, 10 Feb 2009 05:55:01 -0500 (EST)
  • References: <gmp0pi$c02$1@smc.vnet.net>


Hi,

it is common for numerical routines to create such problems. The reason 

is, that a numerical routine can never sample the whole definition 

region. However, it is a bit astonishing that NMinimize alreday fails 

with this small problem. Maybe you want to ask Wolfram what they think 

about this.

Daniel



appris at att.net wrote:

> Here is an example from Mathematica's user's guide:

> 

> In[8]:= NMinimize[x^4 - 3 x^2 + x, x]

> 

> Out[8]=  {-3.513905039, {x -> -1.300839566}}

> 

> however, trying to replicate it, I get the following:

> 

> In[2]:= NMinimize[x^4 - 3 x^2 + x, x]

> 

> Out[2]= {-1.070230182, {x -> 1.130901122}}

> 

> One way to find the global min, I had to use a constraint such as x<0.

> 

> Has anyone come across such a problem?

> 

> In[3]:= $Version

> 

> Out[3]= "7.0 for Microsoft Windows (32-bit) (November 10, 2008)"

> 

> Thanks.

> 




  • Prev by Date: Re: NMinimize Bug in Mathematica 7.0?
  • Next by Date: Re: newbie: explicit function works, "function object" doesn't
  • Previous by thread: Re: NMinimize Bug in Mathematica 7.0?
  • Next by thread: Re: NMinimize Bug in Mathematica 7.0?