[Rivet] CMS analysis of dijet ratios at large rapdities

[Hendrik Hoeth]

Hi Albert,

thanks for the analysis. I've had a look at the code, and I have to
admit that I don't understand the error calculation in your
divide_set_by_subset_with_binomial_errors() function at all. There
is a reference to some ROOT code, but it points to an empty line
(http://root.cern.ch/root/html/src/TH1.cxx.html#2592) -- I can only
suspect that it's meant to point at the TH1::Divide(const TH1 *h1, const
TH1 *h2, Double_t c1, Double_t c2, Option_t *option) in line 2651 in the
current version of TH1.cxx.

Anyway, here is what I find suspicious:

- In the main error formula "sqrt( ((1.-2.*w)*e1*e1 + w*w*e2*e2 )/(b2*b2))"
  the errors appear in fourth power, while in the ROOT version they
  don't.

- Numerator and denominator seem to be swapped compared to the ROOT
  version (b1 = h2.binHeight(i) and b2 = h1.binHeight(i)).

- At the end you divide the error by a factor of 2.

Can you please confirm that this code is actually what you want to do?

Cheers,

    Hendrik

-- 
If your dreams don't scare you, then you are not dreaming big enough.