Location and activity based transmission

[rstager]

Model exposure based on location and activity. Initial thoughts: Slice the day into 1) home, 2) daytime work/school/assited care, 3) other public activities (grocery, gym, church,...). Assign each actor to a location for each time slice (based on actor properties, i.e. age).

Model a more localized transmission. i.e. at work, the actor has a higher risk within a small workgroup, but less risk with people in other workgroups. Also, model a 'super spreader' subgroup at work. i.e. managers interact with workgroup, but also with other managers. (same with students and teachers). Also, model customer-facing employees. i.e. At a grocery store, the cashier has a brief interaction with many customers.

Model different risk activity based on location. i.e. a gym is more risky than a library.

[rstager]

I have started exploring ways to do this. If anyone else wants to collaborate, let me know

[rstager]

In the last discussion, there was a desire to make the location based transmission model more abstract. Let me propose this approach. 1) Every actor is assigned to a 'night' location that is a small group of actors that do not practice social distancing. 2) Every actor is assigned to a 'day' location. Each location has a distribution of group size, and activity risk factor. Selected locations have a distribution of ages-brackets. 3) There are a random number of 'superspreader' actors that are assigned to multiple day locations and act as a link between islands of day locations. 4) Assignment to day locations may be filtered by vaccine status, willingness to do frequent testing, etc. 5) Enhancement to include assignment to additional 'activity' locations for shorter durations to simulate things like shopping, gym, church, etc. Again with a distribution of group sizes and activity risk factors