Re: Problem with fitting

• To: mathgroup at smc.vnet.net
• Subject: [mg101551] Re: Problem with fitting
• From: dh <dh at metrohm.com>
• Date: Fri, 10 Jul 2009 06:42:43 -0400 (EDT)
• References: <h2fe8l\$pvp\$1@smc.vnet.net>

```
Hi Chris,

it looks like your expression does not depend on the parameter.

Try e.g.:

FindFit[Table[i, {i, 8}], x, {{a, 3}}, x]

and you get the message you describe.

Daniel

Chris Schreiber wrote:

> Hi there,

>

> I encounter the following problem when fitting a complicated function using FindFit: when running

>

> BBWorigin =

>  FindFit[Data2, {CHFSorigin[x] /. ABWorigin /. ABW /. BBW   /. CBW /.

>      DBW /. FGW} ,  {{COorigin, 4.3}}, {x}, MaxIterations -> 1000]

>

>

> where I previously defined the function CHFSorigin and where the rules ABWorigin, ABW, BBW, CBW, DBW and FGW were also defined, I obtain the following output by mathematica:

>

> {COorigin -> 4.3}

> FindFit::fmgz: Encountered a gradient that is effectively zero. The \

> result returned may not be a minimum; it may be a maximum or a saddle \

> point. >>

>

> My first reaction was to change COorigin to a different value, but that did not change the result, i.e. the fit always stopped at the initial value. Does anybody have

> any ideas about how to progress from here?

>

> Best wishes, and thanks a lot,

>

> Christian

>

```

• Prev by Date: Re: GeoDistance and transformation from WGS84 to WGS84
• Next by Date: Re: matrix equation
• Previous by thread: Problem with fitting
• Next by thread: Re: Wolfram | Alpha: Acres per square mile? (oddities)