Remember, when you compute frac_susc, you’re asking “how diluted with potential victims is the population that an infectious person encounters when they walk around in society?” Basically, “who’s walking around the mall?” Ask yourself:

• Are susceptible people walking around the mall?
• Are exposed people?
• Are infected people?
• Are vaccinated people?
• Are quarantined people, regardless of how far the disease has progressed in them?
• Are recovered people?
• Are dead people?

Whoever’s walking around the mall is in the denominator. Whoever’s not, is not.