differentiating between baseline and modified contact matrices

[cwhittaker1000]

At the moment, we don't distinguish between the baseline contact matrix (i.e. in the absence of any interventions) and any modification of that contact matrix at the beginning of the model run.

Issue here, that a number of people have been bitten by a problem where if you specify an R0 in run_explicit_SEEIR (e.g. R0 = 3) and then for your contact_matrix_set specify a contact matrix that you've pre-reduced (e.g. to capture interventions in place maybe), then when we do the beta_est_explicit call inside, we end up with a huge beta (because beta needs to be huge to get an R0 of 3 with such a small contact matrix).

This isn't what people are intending and so I propose adding in another argument to run_explicit_SEEIR that'll be "baseline_contact_matrix" which will enable differentiation of the baseline contact matrix and the modified one in the scenario described above!