MathGroup Archive 2006

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

Search the Archive

Optimization doing more function evaluations than claimed

  • To: mathgroup at smc.vnet.net
  • Subject: [mg67333] Optimization doing more function evaluations than claimed
  • From: "Solomon, Joshua" <J.A.Solomon at city.ac.uk>
  • Date: Sun, 18 Jun 2006 05:13:38 -0400 (EDT)
  • Sender: owner-wri-mathgroup at wolfram.com

After months of waiting, I finally got a good reply from WRI. Details below:

On 15/6/06 17:30, "support at wolfram.com" <support at wolfram.com> wrote:

> From: "Solomon, Joshua" Date: Fri, 13 Jan 2006
To: mathgroup at smc.vnet.net
> 
> Dear W.R.T.S.
> 
> Gareth Russell raised this issue on the newsgroup back in May and I echoed it
> last month ( 
> http://groups.google.co.uk/group/comp.soft-sys.math.mathematica/browse_threa
> d/thread/1d6ac336227232aa/cc7e5b6aa1d4a829?lnk=st&q=group%3A*mathematica*+au
> thor%3Asolomon&rnum=1&hl=en#cc7e5b6aa1d4a829 ). No one seems to have replied.
> I was wondering if this were a known bug, or even possibly a feature! For
> reference: In[1]:= {$Version, $ReleaseNumber} Out[1]= {5.1 for Mac
> OS X (October 25, 2004),0}
> -----------------------------------------------------------------------
>
> Mr. Solomon,
> 
> Look at the output from EvaluationMonitor with InputForm. This is probably
> coming from finite difference evaluation of derivatives.
> 
> When I added InputForm wrappers to the Print to see all 16 digits, the small
> variations in the parameters are visible.     Here is an excerpt.
> 
> In[5]:= FindMinimum[err = dummy[a, b, p, pMinusQ], {{a, 10}, {b, 0.03}, {p,
> 3}, {pMinusQ, 0.44}}, EvaluationMonitor :> Print[ "err=", InputForm[err], "
> a=", InputForm[a], " b=", InputForm[b], " p=", InputForm[p], " pMinusQ=",
> InputForm[pMinusQ]]]
> 
> ...
> 
> Bruce Miller Technical Support Wolfram Research, Inc. support at wolfram.com
> 
> ----------------------------------------------------------------------------


  • Prev by Date: Re: Problem with derivate of product
  • Next by Date: Re: Mathematica bugs?
  • Previous by thread: Re: using a previous result
  • Next by thread: Uniform arc length basis curve fitting