Consistent EIR definition in the Pf/Pv models#377
Conversation
RJSheppard
force-pushed
the
match_pv_eir_to_pf_model
branch
from
e9aa687 to
4549107
Compare
giovannic
force-pushed
the
match_pv_eir_to_pf_model
branch
from
4549107 to
9633941
Compare
giovannic
left a comment
giovannic
left a comment
There was a problem hiding this comment.
I feel like this is going in the wrong direction. I think we should be going towards having all EIRs (parameters and outputs) in the same units (perhaps adults per year?). I know we have adults per year for malariaEquilibrium and adults per timestep in the outputs. But I think that's bad.
And now species_eir could represent two different units depending on vivax_EIR_at_population_level? It feels like it's adding to the confusion.
Is there any way to make sure species_eir always represents adult EIR by making changes in compatibility.R?
|
I was trying to adjust MW's model to match the Pf model, while still allowing for use of MW's original EIR definition for comparison if needed. We could decide it isn't important to preserve both methods, though. Completely agree about input/output definitions - that tends to trip up new users! I'm not sure what you mean by the changes in compatibility.R. For the EIR to have the same definition, it just needs to be treated the same way in both models...? |
|
Total population EIRs (as used by the original P. vivax model) can now be input into the set_equilibrium function, which converts to adult EIR. The calculated equilibrium and simulations are then run at the adult level. This produces the matching human prevalence and FOIM outputs, but changes the relationship between EIR/FOIM and total mosquitos in the mosquito equilibrium. The mosquito part of the P. vivax model is not fixed to real mosquito populations and acts as a dial for model calibration, so this adjustment will not fundamentally change the model or model calibration processes. 
|
giovannic
left a comment
giovannic
left a comment
There was a problem hiding this comment.
Will call for more clarification
…tes from EIR were different, where the EIR of falciparum was at the level of the adult, whereas vivax was at the level of the whole population (taking into account fewer bites on children). For consistency, we now remove this difference, opting for the falciparum version. This will not make a difference to outputs, but changes the scaling for EIR to prevalence in the vivax model (EIR is abstract in the vivax model). A similar change is made in the vivax equilibrium package.
…ivax parameter test to account to addition of this parameter.
…urther model functionality to use adult EIR in the vivax model. This reduces differences between the two parasite models and allows semi-backwards compatability where total pop EIR can be input and generate similar human prevalence results to the original model.
giovannic
force-pushed
the
match_pv_eir_to_pf_model
branch
from
5f839cd to
817f417
Compare
|
Merging due to inactivity and approval |
2 participants
Previously, vivax and falciparum calculations to generate expected bites from EIR were different, where the EIR of falciparum was at the level of the adult, whereas vivax was at the level of the whole population (taking into account fewer bites on children). For consistency, we now remove this difference, opting for the falciparum version. This will not make a difference to outputs, but changes the scaling for EIR to prevalence in the vivax model (EIR is abstract in the vivax model). A similar change is made in the vivax equilibrium package.