include FoM-weighting for maps

[tjlane]

Under the assumptions that

1. the structure factor amplitudes are normally distributed
2. the phases are Von Mises (circularly normally) distributed
3. all amplitudes and phases are independent

It is possible to show that the MLE for a difference map is

$\Delta \vec{F} = F_1 e^{i \phi_1} - F_2 e^{i \phi_2}$

Assuming the phases come from a model, $m_1$ is the phase integral estimated by comparing the model phases to $F_1$, and similarly for $m_2$.

Three points:

• this theory needs to be tested empirically
• if we employ a more sophisticated scheme, such as the one recommended by R. Read, if we should weight or not will depend on the final map
• we should be able to re-use code from the ROCKET pipeline to estimate ms