MathGroup Archive 2007

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

Search the Archive

Re: Manipulate with very computationally large problems

  • To: mathgroup at smc.vnet.net
  • Subject: [mg83799] Re: Manipulate with very computationally large problems
  • From: "Fred Klingener" <gigabitbucket at gmail.com>
  • Date: Sat, 1 Dec 2007 05:36:30 -0500 (EST)
  • References: <fiopt2$a5f$1@smc.vnet.net>
  • Reply-to: "Fred Klingener" <gigabitbucket at gmail.com>

"Phil I" <p.ingrey at gmail.com> wrote in message 
news:fiopt2$a5f$1 at smc.vnet.net...
> Hi,
>
> I'm try to use manipulate with a very complicated plot (lots of
> Hypergeometric function) with each plot taking around 15 mins to
> generate. Now when manipulate is used it comes up with $Aborted for
> any but the first plot. Mathematica seems to try and calculate each
> plot as I move the control bar, is there a way to get mathematica to
> pre-render all the data first then output the result?

Manipulate is way too cool to limit it to the things that Wolfram intended 
it to do.

First, read or reread the Advanced Manipulate Functionality tutorial. (I 
have to constantly reread all four of the Manipulate and Dynamic tutorials, 
getting something new each time.) Look for what ContinuousAction, 
ControlActive, or TrackedSymbols will do for you. Understand synchronous and 
asynchronous processes and how to control them.

Second, depending on your application - the spans and continuity of 
interactive elements and the amount of computation - you might be able to 
exploit Interpolations. Pre-rendering? Probably not, but stuffing some 
Interpolations into Manipulate's Initialization block can help out a lot. 
Depends though.

The default Manipulate has some restrictions on compute time, and if you 
have ambitions to make a Demonstration, that'll have some more.

Don't give up.

Hth,
Fred Klingener 



  • Prev by Date: Re: Contour levels
  • Next by Date: Re: Simplifying an expression - with my own definition of what "simple"
  • Previous by thread: Re: Contour levels
  • Next by thread: Re: Simplifying an expression - with my own definition of what "simple"