Currently, we use attributes based on class names to access variables, e.g. sim.diseases.sir.susceptible. While we definitely want to keep this as an option, it's annoying that this then requires so many things to be hard-coded, eg ss.sir_vaccine(). This also breaks, I think, if you give a disease a different name. Need to think about what the most robust / least surprising way of handling this would be.
As a trivial workaround, I have been passing in a list of one or more disease class name strings, e.g. sir to my interventions. Works great, but may not be the pattern we choose to persist.
Currently, we use attributes based on class names to access variables, e.g.
sim.diseases.sir.susceptible
. While we definitely want to keep this as an option, it's annoying that this then requires so many things to be hard-coded, egss.sir_vaccine()
. This also breaks, I think, if you give a disease a different name. Need to think about what the most robust / least surprising way of handling this would be.