search

trvrb / antigen

Simulating virus evolution and epidemiology
http://bedford.io/projects/antigen/
19 stars 8 forks source link

Allow halting conditions #6

Open trvrb opened 10 years ago

trvrb commented 10 years ago

Each timestep look at genetic diversity, quit simulation if greater than some number.

trvrb commented 10 years ago

@sarahCobey do we really need this? I would think explosive diversity would be a proper model output under some regimes.

sarahCobey commented 10 years ago

I like to think of it as a cutoff for what constitutes 'explosive.' In my experience, running the sim under this regime can take an extremely long time if the diversity continues to expand. Your call, though.

On Sunday, February 16, 2014, Trevor Bedford notifications@github.com wrote:

@sarahCobey https://github.com/sarahCobey do we really need this? I would think explosive diversity would be a proper model output under some regimes.

Reply to this email directly or view it on GitHubhttps://github.com/trvrb/antigen/issues/6#issuecomment-35212569 .

trvrb commented 10 years ago

@sarahCobey @OanaCarja I'd like to implement this, but need a good definition of explosive. Do we just look at genealogical diversity? Do we need to consider antigenic diversity as well? Two branches (a single bifurcation event) will have very high genealogical diversity, but should be alright from a computational standpoint.

sarahCobey commented 10 years ago

Shouldn't genealogical diversity be enough? Say, if neTau is extremely high after some initial burn-in period? (I've noticed it can spike initially in flu-like simulations.)

On Tue, May 20, 2014 at 4:49 PM, Trevor Bedford notifications@github.comwrote:

@sarahCobey https://github.com/sarahCobey @OanaCarjahttps://github.com/OanaCarjaI'd like to implement this, but need a good definition of explosive. Do we just look at genealogical diversity? Do we need to consider antigenic diversity as well? Two branches (a single bifurcation event) will have very high genealogical diversity, but should be alright from a computational standpoint.

— Reply to this email directly or view it on GitHubhttps://github.com/trvrb/antigen/issues/6#issuecomment-43687956 .

trvrb commented 9 years ago

Attack rate? Time profiling? Memory profiling?