
1. create input histograms for simple counting experiment.
root -b -q makeHistograms.C

see makeHistograms.C for observed data value, and signal, background expectation values.
Error values are set in: config/example_channel.xml

2. create workspace using RooStats histfactory machinery.
hist2workspace example.xml

--> By default there is an uncertainty on the background estimate, which is turned on.
The x-section uncertainty and luminosity uncertainties of the signal 
estimate are included in the construction of the pdf, but have been turned off in the pdf
(set to constant in example.xml)

3.
Running this last command creates several workspaces, with different error treatments:
Gauss, log-normal, Gamma, as configured in example.xml.

Also, this automatically does the fit of the fit to data, and create profile
log likelihood plots. (Option can be turned off with: ExportOnly="True",
in example.xml)


