Services & Resources / Wolfram Forums
-----
 /
MathGroup Archive
1995
*January
*February
*March
*April
*May
*June
*July
*August
*September
*October
*November
*December
*Archive Index
*Ask about this page
*Print this page
*Give us feedback
*Sign up for the Wolfram Insider

MathGroup Archive 1995

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

Search the Archive

Re:Re: Clear value of pre calculated expresion

  • To: mathgroup at christensen.cybernetics.net
  • Subject: [mg1702] Re:[mg1491] Re: [mg1438] Clear value of pre calculated expresion
  • From: mvyver at ecel.uwa.edu.au (Mark Van de Vyver)
  • Date: Mon, 17 Jul 1995 03:54:30 -0400

Hi Mathgroup,

When running some simulations in mma2.2 under WFWorkgroups 3.11 and/or unix
I find that the memory (real and virtual) fills up....for large sims this
happens quite fast. Clear[] and Remove[] provided no joy. Since I'm using
some 3rd party packages I was ready to put the blame on them and postponed a
fix. Now, having seen the msgs below I'm wondering if the problem is that
mma is remembering the prior sim. results and calculations. Is this
possible? does anyone have anything to add to what Dave Wagner says (ie this
caching can't be overcome)?
The sim is in the following form (there is no screen output):

K=20
Do[ function[k]'s;
Write[]'s;
clear[]'s;
etc, { k, K} ]

Thanks for any help.

Cheers

Mark Van de Vyver

THE INCLUDED MESSAGES:

In article <3rqk54$hs3 at news0.cybernetics.net>,
John Fultz  <jfultz at wri.com> wrote:
>> I would like to clear value of pre calculated expresion which Mathematica
>> save in memory. When you do 
>> 
>>              N[Pi,23]
>> 
>> Mathematica would calculated number Pi on 23 places and save that value in
>> work memory for future calculations, but some times ( test and simulation )
>> is nice to dont that ... so my question is how clear that type of data from
>> working memory ? 
>> 
>> Dubravko Penezic
>
>You can do this by evaluating the following lines:
>
>Unprotect[In,Out]
>Clear[In,Out]
>
>This will clear the whole command history up to the current evaluation.
>
>John Fultz
>Wolfram Research, Inc.

I don't think that's what Mr. Penezic was talking about.  The kernel
caches the numerical values of things that have been evaluated for some
time, e.g.:

    In[1]:=
	N[Pi,1000]; // Timing
    Out[1]=
	{0.216667 Second, Null}

    In[2]:=
	N[Pi,1000]; // Timing
    Out[2]=
	{0. Second, Null}

At some point in the distant future, this computation will require .217
seconds again, but there's no (known) way to predict when or to force it to
forget the cached answer.  Clearing In/Out has no effect on this.

Except for benchmarking Mathematica itself, I'm not sure why one would
 want to clear the cached values, but I think that's the question
being asked.

		Dave Wagner
		Principia Consulting
		(303) 786-8371
		dbwagner at princon.com
		http://www.princon.com/princon
Mark Van de Vyver

#############################################################################
#                                                                           #
#  Department of Accounting and Finance      Phone: (61) (09) 380-2510      #
#  University of Western Australia           Fax: (61) (09) 380-1047        #
#  Nedlands   6907                           e-mail: mvyver at ecel.uwa.edu.au #
#  Perth                                                                    #
#                                                                           #
#############################################################################





  • Prev by Date: How to apply EdgeForm to ParametricPlot3D?
  • Next by Date: Re: How do I find my current directory?
  • Previous by thread: Re: How to apply EdgeForm to ParametricPlot3D?
  • Next by thread: Windows front end & remote kernel