Note: This section is currently under construction - check back soon for updates!
In the real world, a typical and sufficiently mature analysis will inevitably contain many components (a complex cutflow, systematic uncertainties, data-driven backgrounds, hundreds if not thousands of samples, etc.) which can quite often serve to impede the learning process of its underlying framework.
In contrast to this picture, what follows is an example which demonstrates a bare minimum analysis using CAF, stripped of as many elements as possible. Using a single (BDT trained) VBF signal Monte Carlo sample as input, a simple 2 step cutflow is defined and a single histogram produced. This page follows the "Hello World!" philosophy used in computer programming - that is, although the example is admittedly not so realistic, something simple you can work with is always better than lots of detail you can't actually implement.
Please follow the top level CAFExample README at least through the 'build the project' section in order to set up a working directory for the following exercise.
Prepare step (prepare.py)
Initialize step (initialize.py)
Analyze step (analyze.py)