Re: plot calculated IR-spectrum

On Jul 3, 1:15=A0pm, Philipp Eiden <pei...@[EMAIL PROTECTED]
> wrote:
> Hi everybody,
>
> I would like to plot an IR-spectrum. The problem is,
> that I have only the datapoints of the intensity and
> wavenumber at the maximum, e.g.
>
> # spectrum.dat
> # v =A0 =A0 I =A0 =A0 =A0
> 0.00 =A0 =A00.00
> 108.97 =A00.00
> 169.61 =A06.21
> 325.48 =A00.00
> 486.67 =A0183.07
>
> so I would need to fit a Lorentzian function to each
> datapoint-pair
>
> f(x) =3D alpha*I_(1strow)/(pi*alpha*(x-v_(1strow))**2+1)
> g(x) =3D alpha*I_(2ndrow)/(pi*alpha*(x-v_(2ndrow))**2+1)
> .
> .
> .
> z(x) =3D alpha*I_(last-row)/(pi*alpha*(x-v_(last-row))**2+1)
>
> and then sum all functions to obtain the spectrum. Is
> this possible with gnuplot? How?
>
> Thanks
>
> Philipp Eiden

I don't have time to think this through completely but one way would
be call an awk script that parsed each line into gnuplot set
commands , eg:

v_2 =3D169.61; l_2 =3D6.21

or more directly, your function definitions. Call the script from
gnuplot (help system) then read the output into gnuplot with load
command and plot f(x)+g(x)....+z(x)

$0 is the line counter for gnuplot data but I can't think of a clean
way to use that to be part of a varable name.

You will obviously need to make sure your data are valid : (0,0) ?!!