Re: [ROOT] log-scale and TH1::Fit

From: Thomas Bretz (tbretz@uni-sw.gwdg.de)
Date: Mon Oct 29 2001 - 16:47:35 MET


Hi Rene,

> Well, obviously, you should not fit with a "gaussian" what is not a gaussian.
> You should define a function with the right variable.

It is. The logarithmic data is a gaussian. But it is more convinient to
display the raw data with a logarithmic scale (10^3, 10^4, instead of 3,
4, ...) like in your example.
(For interest: I'm talking about some data vs. energy. It is more
conviniet to have a logarithmic x-axis with labels 10^3[E/GeV],
10^4[E/GeV] for the display, because it is easier to read and
understand, than 3[log10(E/GeV)], 4[log10(E/Gev)]. But data vs.
log10(energy) is gaussian. Now I dislike the idea of writing an new
function TF1 of a gaussian which is already implemented, only to add the
additional log10() at the right place in the code)

> What I could do automatically is the creation of the histogram with
> the variable size bins computed with the logarithm.

Nevertheless: That's a good idea...

Best regards,
Thomas.



This archive was generated by hypermail 2b29 : Tue Jan 01 2002 - 17:51:05 MET