Config parser

[aguinot]

Build a config parser to handle the catalogues and the processing.
The config file use the Yaml format with custom features.
See PR #2

Note: Here all the "processing" (the call the eval fonction) needs to be lazy and will be done when the catalogs are loaded.

Features:

• Read .yaml files
• Setup workspace (maybe not during the configuration step)
• Take care of necessary inputs for catalogs (ra, dec, e1, e2, ...)
  • deal with Lensfit like catalogs
  • deal with Metacal like catalog
• Handle star catalog
  • Make sure the catalog is consistant with the galaxy catalog
• Handle multiples catalogs
  • and keep track of which entries belongs to which catalogs
• Possibility to creates new columns based on existing one based on basic operations (using python built-in eval fonction)
• Possibility to define variables in the config file and use them to give more flexibility (especially to handle multiple catalogs)
• Handle all the plotting functions
  • plot residuals (see plot residuals #5)
• Have a separate config for cluster computation
• Have a separate config for matplotlib parameters
• Make it possible to skip the computation and just re-do the plot
• Make it possible to combine plots from several runs on different catalogs

[aguinot]

At the moment it is possible to work on an already existing workspace directory. This is useful if you want to add a plot without having to re-do all the computation. By default the code will not recompute something that already exist. Not sure if this is the best way to do.