       Re: Help with fitting

• To: mathgroup at smc.vnet.net
• Subject: [mg131495] Re: Help with fitting
• From: Bob Hanlon <hanlonr357 at gmail.com>
• Date: Sat, 10 Aug 2013 04:38:54 -0400 (EDT)
• Delivered-to: l-mathgroup@mail-archive0.wolfram.com
• Delivered-to: l-mathgroup@wolfram.com
• Delivered-to: mathgroup-outx@smc.vnet.net
• Delivered-to: mathgroup-newsendx@smc.vnet.net
• References: <20130809054847.0CB036A06@smc.vnet.net>

```data = {{0.1, 1400}, {0.5, 10}, {1, 5}, {5, 20}, {10, 250}};

I am guessing that this is what you intended for the equation

eqn = Log [y/b] == Log [a*x/b + a*x/b] + (-0.22 + Log [a*x/b]^2) //
Rationalize;

expr = (y /. Solve[eqn, y][])[];

param = FindFit[data, expr, {a, b}, x]

{a -> 1.14522, b -> 2.27658}

LogLogPlot[
Evaluate[expr /. param],
{x, .05, 20},
Epilog -> {Red,
AbsolutePointSize,
Point[Log[data]]}]

Bob Hanlon

On Fri, Aug 9, 2013 at 1:48 AM, ismail <ism45 at yahoo.com> wrote:

> Hello,
>
> I have experimental data with (y) and (x)
>
> I need to fit them into the following model
>
> Log (y/b) = Log (ax /b/1+ax/b) + (-0.22/1+(log ax/b)^2)
>
> The result should be a plot with Log y on the y-axis, Log x in the x axis.
> and the two constants a and b will be the result of the fitting.
>
> I tried that in Mathematica, but my problem is that I don't know how to
> define the function so I can keep (Logy/b) in the left side.
>
> Mathematica allows me only to defin a model in the form (y=f(x))
>