Script to perform a PrEW
fit using the PrEWUtils
library.
- Make sure you installed
PrEW
andPrEWUtils
(preferably in a directory next to this one). - Make sure your software is up-to-date (ROOT, gcc, cmake, ...) or load using macro (only on NAF):
cd macros && source load_env.sh && cd ..
- Compile the code:
cd macros && chmod u+x compile.sh && ./compile.sh && cd ..
... simply recompile it using the macros/compile.sh
macro.
The code can be easily run after compilation:
cd bin && ./PrEWRunExample && cd ..
This will produce an output file in the output
directory.
It is a plain text file that stores some information about the results of the fits
The executable can also be run on the BIRD cluster using the HTCondor scheduler.
cd macros && ./run_on_cluster.sh [--cpus=n_cpus] && cd ..
With the --cpus=n_cpus
the number of cpus used on the BIRD cluster can be set.
The source code is in source/main.cpp
.
It uses the necessary PrEW
classes and the interfaces provided by PrEWUtils
to be as clear as manageable.
(This may not always have worked out.)
For open questions please consult the PrEW
and PrEWUtils
source code or open an issue on the GitHub page.
A jupyter (python) notebook is provided to interpret the output of the toy fits.
It sits in the notebook
directory and can be opened using
cd notebook && jupyter notebook ResultNotebook.ipynb && cd ..
and will produce histograms in the output
directory.
The notebook can also be run without opening a browser window using
cd notebook && jupyter nbconvert --to notebook --inplace --execute ResultNotebook.ipynb
Text outputs (which would normally be visible in the browser window) can be found in the ResultNotebook.ipynb
file.