Re: how to start Solve[] with an initial solution guess?
- To: mathgroup@smc.vnet.net
- Subject: [mg12490] Re: how to start Solve[] with an initial solution guess?
- From: "Allan Hayes" <hay@haystack.demon.cc.uk>
- Date: Tue, 19 May 1998 13:31:53 -0400
- References: <6jf6bv$4ai@smc.vnet.net>
Bill Christens-Barry wrote in message <6jf6bv$4ai@smc.vnet.net>...
>I have a system of equations that Mathematica 3.0 bogs down on for most
>problems of interest. I would like to seed a solution that is found
>for one set of parameters into the routine for finding the solution of
>a "nearby" problem (meaning one with only slightly different parameter
>values).
>
>How can I do this? I have tried including a set of conditions (bounds)
>on the solution space as inequalities in the system of equations to be
>solved, but Mathematica gagged on the form of the inequalities.
>
>Are there other methods of launching Solve[] from an initial trial
>solution?
>
>Should I be trying to learn how to instruct Solve[] on what methods it
>should be using? If so, where can I learn about this?
>
>Examples that illustrate any explanation would be especially helpful.
>
A quick response -will need some progamming to get it automatic
eqns = {x+5 == 3};
soln =Solve[eqns,x]
{{x -> -2}}
nearsoln1= {{x->-2 +dx}}[[1]];
{x+4==5}/.nearsoln
{2+dx==5}
Solve[%,dx]
{{dx->3}}
Allan Hayes
Mathematica Training and Consulting
Leicester UK
http://www.haystack.demon.co.uk
hay@haystack.demon.co.uk
voice: +44 (0)116 271 4198
fax: +44(0)116 271 8642