Multiclosure data keyword
Created by: wilsonmr
I think after this, the new runcards can be used with all the closure actions. I have fixed that the grouping is always done on the experiment level. In the future when one wants to do closures for theory covmat fits then somebody will have to add some more structure to the totals code (similar to #930 ), this seemed superfluous right now since at that stage one will also have to change the loader functions and by then I hope we will have the superior python interface
I ran a report with all of the actions from multiclosure which would have been affected with both an old runcard:
https://vp.nnpdf.science/JckSVWl7QACJrcChhq8ZKg==
and a new one: