differential waning of infection and vaccine derived NAT

[slwu89]

from Azra:

First step would be to track NATs separately for vaccination and natural immunity - and then also track the sum of these so that they can be used jointly to estimate vaccine efficacy.

Second would be to allow different parameters (the durations of short / long and the timeswitches) for natural immunity, dose 1/2 of the vaccine and dose 3 of the vaccine (where dose 3 parameters would replace dose 1/2.

That would give us maximum flexibility in fitting models to the observed data in the coming week or two to see if we can see a slower decay after dose 3 and/or from natural immunity (which would represent broader immune responses / T cells). At the moment the only way we can vary this is by the boost that is a separate parameter for each dose and for natural immunity.

[slwu89]

New parameters:

1. dr_vec_doses is a matrix where rows are days and columns are each dose.
2. dr_vec_inf is a vector where rows are days
3. max_ab_inf maximum value for infection derived NAT

New variables:

1. ab_titre_inf a DoubleVariable object

[slwu89]

New functions:

1. R/variables_nat.R: create_independent_nat_variables
2. R/process_nat.R: independent_ab_titre_process, get_time_since_last_infection
3. R/events_nat.R: attach_event_listeners_independent_nat
4. R/efficacy_nat.R: make_calculate_nat
5. R/parameters_nat.R: make_independent_vaccine_infection_nat_parameters

Modified functions:

1. R/process_infection_vaccine.R: infection_process_vaccine, infection_process_vaccine_cpp
2. src/process_infection_vaccine.cpp: infection_process_vaccine_cpp_internal
3. R/events_exposure_vaccine.R: create_exposure_scheduler_listener_vaccine

[slwu89]

done #65