mrc-ide / squire

SEIR transmission model of COVID-19. Documentation at:
https://mrc-ide.github.io/squire/
Other
50 stars 28 forks source link

Baseline_contact_matrix argument addition #114

Closed cwhittaker1000 closed 4 years ago

cwhittaker1000 commented 4 years ago

This PR introduces the baseline_contact_matrix argument into run_explicit_SEEIR and run_deterministic_SEEIR. Specifically, it:

Note: I've noticed that calibrate already has an argument called baseline_contact_matrix. @OJWatson have started from a position of being cautious and so haven't touched that yet. Would be good to get your thoughts on how to harmonise the two - won't be difficult as they're doing the same thing - just don't want to touch calibrate until we've spoken in case I break anything!

Context This has been introduced as a number of users were getting caught out when putting in an already reduced contact matrix in as the argument in contact_matrix_set (to e.g. simulate the effect of the epidemic starting with control measures already in place). Because contact_matrix_set was being used used to calculate beta via beta_est_explicit, the result was no reduction in R0 and instead beta_est_explicit would just produce a higher beta to give the same R0 for this reduced contact matrix. Introducing baseline_contact_matrix as an argument to the run functions, and ensuring this is passed to beta_est_explicit, avoids this problem!