Services & ResourcesWolfram ForumsMathGroup Archive

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

Re: Re: Gaussian intersection

• To: mathgroup at smc.vnet.net
• Subject: [mg39143] Re: Re: Gaussian intersection
• From: "Michal Kvasnicka" <michal.kvasnicka at quick.cz>
• Date: Fri, 31 Jan 2003 04:36:56 -0500 (EST)
• References: <b10e8t$mr0$1@smc.vnet.net> <b15omg$et0$1@smc.vnet.net> <200301290835.DAA21127@smc.vnet.net> <b1aga8$pi7$1@smc.vnet.net>
• Sender: owner-wri-mathgroup at wolfram.com

For s1 = s2 you must obtain x = 1/2(m1+m2), due to the symmetry of the
problem, but your formulation gives "nothing".
The Gaussian normal distribution has the following form:
1/(Sqrt[2Pi]*sigma)*Exp[-(x-mu)^2/(2*sigma^2)]. Am I right?

Michal

"Dr Bob" <drbob at bigfoot.com> pí¹e v diskusním pøíspìvku
news:b1aga8$pi7$1 at smc.vnet.net...
> Better yet,
>
> Off[Solve::ifun]
> First@Solve[Exp[-(x - m1)/(2*s1)]/Sqrt[s1] == Exp[-(x - m2)/(2*s2)
> ]/Sqrt[s2], x];
> PowerExpand[% /. {s1 -> r^2, s2 -> s^2}]
>
> Bobby
>
> On Wed, 29 Jan 2003 03:35:35 -0500 (EST), Michal Kvasnicka
> <michal.kvasnicka at quick_nospam.cz> wrote:
>
> > Or better:
> >
> > Solve[Exp[-(x - m1)^2/(2s1^2)]/s1 == Exp[-(x - m2)^2/(2s2^2)]/s2, x]
> >
> > Michal
> > "Jens-Peer Kuska" <kuska at informatik.uni-leipzig.de> pí¹e v diskusním
> > pøíspìvku news:b15omg$et0$1 at smc.vnet.net...
> >> Hi,
> >>
> >> Solve[Exp[-(x - m1)/(2s1)]/Sqrt[s1] == Exp[-(x - m2)/(2s2)]/Sqrt[s2],
x]
> >>
> >> ??
> >>
> >> Regards
> >> Jens
> >> Vaidyanathan wrote:
> >> >
> >> > Can anyone please tell me how to find the intersection of two
> >> gaussians?
> >> > Is there any standard method to do that?
> >> > Thanks,
> >> > Vaidyanathan.
> >> >
> >> > --
> >> > Vaidyanathan Ramadurai
> >> > Graduate Student
> >> > http://www4.ncsu.edu/~vramadu
> >>
> >
> >
> >
> >
>
>
>
> --
> majort at cox-internet.com
> Bobby R. Treat
>
>

• References:
  • Re: Gaussian intersection
    • From: "Michal Kvasnicka" <michal.kvasnicka@quick_nospam.cz>