Re: FindFit issue

*To*: mathgroup at smc.vnet.net*Subject*: [mg86406] Re: FindFit issue*From*: Antti Penttilä@smc.vnet.net*Date*: Tue, 11 Mar 2008 02:53:55 -0500 (EST)*Organization*: University of Helsinki*References*: <fr2mhj$o5h$1@smc.vnet.net>

Hi, With a little help FindFit will find the right solution: FindFit[d, {0.072214 - a*Log[10, 1 + b*x], b > 0}, {{a, 0.05}, {b, 0.01}}, x] return {a -> 0.0340109, b -> 0.00568993} With a nonlinear model it's probably best if you try first to bracket the parameter values near the optimal by e.g. plotting the solution and the data and then give these parameter values as a starting parameters for FindFit. There are a couple of different optimization routines in Mathematica. NMinimize is probably better in finding the global optimum. If you write out the squared residual sum between your model and data as a function (mfun in my example below), and minimize that you get the right solution without specifying starting values: NMinimize[ {mfun, b > 0}, {a, b}] return {9.81295*10^-7, {a -> 0.0337413, b -> 0.00578092}} Antti t.dubaj at gmail.com wrote: > Hi, > I am trying to fit experimental data > > d = {{1000, 44.576*10^(-3)}, > {600, 49.546*10^(-3)}, > {800, 46.757*10^(-3)}, > {400, 55.023*10^(-3)}, > {200, 61.209*10^(-3)}, > {10, 71.584*10^(-3)}, > {100, 65.367*10^(-3)}} > > and model is: > > y = 0.072214 - a*Log[10, 1+b*x] > > but Mathematica after typing > > FindFit[d, 0.072214 - a*Log[10, 1 + b*x], {a, b}, x] > > return > > FindFit::nrlnum: The function value {-0.0226973-0.0208168 \ > \[ImaginaryI],<<5>>,-0.0282009-<<21>> \[ImaginaryI]} is not a list of > \ > real numbers with dimensions {7} at {a,b} = {0.0152574,-1.99204}. > > {a -> 0.0152574, b -> -1.99204} > > (especialy value of "b" is really awful - negative number in Log) > > But I know, there is Fit (calculated in Scientist) with coefficients > a = 0.033706 > b = 0.0057934 > > Can someone explain what I'm doing wrong? > > Best regards, > > T.D. > > >