[WIP]: Point prescriptions

Created by: RosalynLP

I've been working on implementing a 3 pt covmat as detailed in Luca's note, as this is the simplest example of the different point prescriptions.

So far it runs but I don't trust the results I'm getting so I need to look into it a bit more.

The way the code roughly works is that I separated everything out into different processes, applied Luca's prescription between each pair of processes separately to give N_proc^2 sub-matrices, and then combined these through keeping track of indexing to make a total covariance matrix, ordered by process. I then applied a mapping to map points in this matrix onto a covariance matrix ordered by experiment, as they appear in the runcard. This can then be combined with experimental covariance matrices by direct addition.

The output is:

which doesn't make any obvious sense to me. However, I think I may have not yet implemented the prescription correctly, in that I should be applying a different prescription to the block diagonal elements (i.e. with different normalisation when the procedure is applied between two regions with the same process type). I will work on making this change and see where it leads, but though I'd open the PR for now.

One other thing is I branched off proc2 which hasn't yet been merged. I also had to comment out the covariance matrix function from proc2 just so I didn't have to define a whole bunch of new plotting funtions etc. in order to see the output.