Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

WIP: modeling #1

Open
wants to merge 42 commits into
base: master
Choose a base branch
from
Open

WIP: modeling #1

wants to merge 42 commits into from

Conversation

beansrowning
Copy link
Owner

@beansrowning beansrowning commented Jul 26, 2019

What

Found some decent code from the West Africa Ebola outbreak which was use for simulations and parameter estimations that I could retool for the current DRC outbreak.

TODO

  • Re-work code from King, et al. to work on HPC and desktop
  • Summarize regional data to national, so we can work with known denominator data
  • Approximate R0, different parameter estimates, along with epidemic trajectory

@beansrowning beansrowning self-assigned this Aug 1, 2019
@beansrowning
Copy link
Owner Author

Some preliminary results from attempting to estimate R0

image

@beansrowning
Copy link
Owner Author

image

  • In both models, R0 seems to be ~ 1.1

image

  • Raw counts seem to outperform cumulative counts in looking at the log-likelihood surface

@beansrowning
Copy link
Owner Author

beansrowning commented Aug 15, 2019

Most of the code has been translated/updated.

Issues now with incoming data breaking counts (#2).

R0, k estimation script tested working on Linux (FORK, MPI) and Win 10 (PSOCK).

Seeming lack of scaling on MPI compared to FORK, but may be due to internal setup within the CDC HPC.

@beansrowning
Copy link
Owner Author

Status

Was working on the forecasting end but that's been stalled for the time being.

Finally found some regional denomonator data (thanks @cliu822).

Now thinking of ways to extend this mass action model to include variable interaction between regions where reliable mixing rates are not known apriori.

A good starting place might just be looking at queen neighbors and computing a naive weighted distance and construct a WAIFW matrix from there. 🤷‍♂

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant