Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.graphics.apps.gnuplot > #4170

Re: different behaviour gnuplot4 and 5 concerning "Implied independent variable y not found in fit function"

Show all headers | View raw

On Thursday, March 21, 2019 at 10:02:24 PM UTC+1, Karl Ratzsch wrote:
> Am 21.03.2019 um 16:02 schrieb hugocoolens@gmail.com:
> > In a batch-file which worked fine for Gnuplot 4, I had the following lines:
> > f(x) = a*10e-8*(exp(x/b)-1)
> > err1(y)=0.01*0.033*y+2e-9
> > err2(y)=0.01*0.0031*y+20e-9
> > err3(y)=0.01*0.0034*y+200e-9
> > err4(y)=0.01*0.0045*y+2e-6
> > error(y)=(y<10e-6)? err1(y):(y<100e-6)? err2(y):(y<1e-3)? err3(y):err4(y)
> > a=30e-3
> > b=50e-3
> > fit f(x) '1n4148_linstep_1na_10ma.dat' using ($1):($2):(error($2)) via a,b
> > 
> > This worked fine in Gnuplot 4, however when I try this in Gnuplot 5, I get the following error message:
> > 	> Implied independent variable y not found in fit function.
> > 	> Assuming version 4 syntax with zerror in column 3 but no zerror keyword.
> > 
> 
> It's only a warning, not an error. Doesn't the fit converge?
> 
> I tried this:
> 
>    set xr [0:1e-3]
>    fit f(x) '+' using ($1):(f($1)):(error(f($1))) via a,b
> 
> where gp of course can't compute errors from the perfect data, but
> otherwise works fine.
> 
> To get rid of the warning, you specify to which variable the
> additional error column refers:
> 
>       fit f(x) '1n4148_linstep_1na_10ma.dat' using
> ($1):($2):(error($2)) yerror via a,b
> 
> (you can also give "zerror" or "error z", that's all the same if you
> only have one dependent and one independent variable.)
> 
> I understand the warning to be just one warning, not two: You gave
> three columns, but the fit function only has one variable. That
> makes gnuplot assume (correctly,) that you don't have two
> independent variables, but one, plus an additional column with
> y(==z)-errors.
> 
> 
>    Best, Karl

Thanks for the help offered. The fit did indeed still converge and the warning disappeared by adding yerror before via a,b 

kind regards,
Hugo

Back to comp.graphics.apps.gnuplot | Previous | Next — Previous in thread | Find similar

Thread

different behaviour gnuplot4 and 5 concerning  "Implied independent variable y not found in fit function" "hugocoolens@gmail.com" <hugocoolens@gmail.com> - 2019-03-21 08:02 -0700
  Re: different behaviour gnuplot4 and 5 concerning "Implied independent variable y not found in fit function" Karl Ratzsch <mail.kfr@gmx.net> - 2019-03-21 22:02 +0100
    Re: different behaviour gnuplot4 and 5 concerning "Implied independent variable y not found in fit function" "hugocoolens@gmail.com" <hugocoolens@gmail.com> - 2019-03-22 01:14 -0700

csiph-web