Documentation & Installation changes + Preshower Correction

[maxgalli]

Since there are chances that neither of us will still be in the loop when this will have to be run once again for Run 3 data, I took the liberty of making some (I hope) useful changes in terms of installation/structure and documentation.
This should help the next poor soul that has to run them to go smoothly through the whole process by him/herself, without the need of asking us (as I said above, I think we should tackle the worst case scenario in which at that time we won't be in the loop anymore).

As you can see there are plenty of commits, but the actual changes are very few and none of them affects the internal behavior.

They can be mostly summarized in what follows:

• refactoring in a Python-package friendly style; as explained in the documentation, this allows a simple installation with python setup.py install and makes it possible to import the package from every location, without the need to update the PYTHONPATH;
• porting to Python3 of the whole package;
• add Ray as a second option for the internal parallelization backend (i.e. joblib is still used everywhere and setting IPyParallel is as well still possible); I have a branch where I do the same with Dask, but hitting this problem so often made me think that it's not a good idea to include it as an option (for now);
• refactor the folder with code for execution by dividing it into the steps that must be followed (check the folders 1-9 inside utils);
• instructions and documentation for every step of the process (see the READMEs inside the above mentioned 1-9 folders).

N.B.: I kept the old code under the legacy folder; it can of course still be run with the changes I made, but the imports have to be changed (and also the name of the setupJoblib function, see 76a2ed9)

Update 13/05/2021: a new section (with documentation) was added for the preshower correction.