Feature small covariance fixes

[henry-wallace-phys]

Two small tweaks.

• I noticed oscillation parameters weren't being given the correct priors so I've just made the prior-setter fully explicit
• delta-cp could, hypothetically be thrown outside of [-2 pi, 2pi] and not be set correctly. I've used a bit of modular arithmetic to prevent this.

Prior Setting :

dcp stays within the correct bounds

dcp_trace.pdf