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

Re: NDSolve - how to bypass safety chceck?

• To: mathgroup at smc.vnet.net
• Subject: [mg111409] Re: NDSolve - how to bypass safety chceck?
• From: "slawek" <slawek at host.pl>
• Date: Sat, 31 Jul 2010 02:41:09 -0400 (EDT)
• References: <i2ub5q$jfj$1@smc.vnet.net>

U=BFytkownik "Daniel Lichtblau" <danl at wolfram.com> napisa=B3 w wiadomo=B6ci grup
dyskusyjnych:i2ub5q$jfj$1 at smc.vnet.net...
> (2) Jettison the DDE approach, and use recurrences to approximate the
> behavior. This is a way of emulating a blind fixed step approach (that
> is to say, there will be no error assessment, so the noise is allowed to
> stay.

Yes, yes, yes.


Nevertheless, you can see, that there is no "an easy option" in NDSolve to
switch off the "safety check", like:

 dirtysolution = NDSolve[...,
BypassAllChecksAnGiveAnyAndPossibleWrongAnswer->True]

or

 evenmorestupid = NDSolve[..., UseDumbFixedEulerMethod->True]


Obviously, it would be funny to write an homemade DumbNDSolve[...] myself.

The idea is to use

Check[NDSolve[...], DumbNDSolve[...]]//Quiet

instead of

Check[NDSolve[...],"Unstable"]//Quiet

slawek