Project Goals and Progress

[ChristyLAI99]

✅ Functions already implemented

• simulate_sirs_det.R– deterministic SIRS simulator (single population)
• simulate_sirs_stoch.R – stochastic SIRS simulator
• simulate_sirs_multi_det.R – deterministic multi-population simulator
• make_beta.R – helper to generate constant or seasonal β
• make_contact.R – helper to generate contact matrices with within-group mixing
• plot_sir_diag.R – plotting for single-pop SIR/Incidence
• plot_multi.R – plotting for multi-pop outputs
• summarize_sim.R – summarise peak infections, incidence, and final R
• demo_run.R, demo_multipop.R – example scripts

🚀 Next steps

• Open an issue for development goals
• Build and maintain a test/demo script that runs through all functions consistently
• Extend stochastic simulations to multi-population setting

🎯 Future goals

• Add documentation & examples in README and vignette
• Decide on design choice:
  - One unified function with a stochastic = TRUE/FALSE toggle
  - Or separate functions (det vs stoch)
• Explore visualisation improvements
• Pack up as an R package
• Design an R package logo (if needed)

🔮 Other possible goals

• Forecasting scenarios
• Consider additional compartments
• Parameter estimation / model fitting (link simulations to real data)
• Sensitivity analysis (robustness testing)
• Scenario comparisons (e.g. different interventions)
• Interactive visualisation (e.g. Shiny App)
• Add a “play/run” wrapper function that asks the user key inputs and automatically runs + plots the simulation

📝Uni Task

• Presentation Oct 20 / Oct 21
• Report Nov 3 (TBC)

[AugustHao]

well done guys, I had a run on the demo and the test scripts and they look good. I suggest tackling the following over the next 2 weeks:

• use a fixed beta + stochastic SIRS simulation to explore behaviour: I noticed the simulation run might be diverging from mean behaviour a bit too much. I would fix beta to sanity check.
• develop plotting code to overlay deterministic simulation as lines, plotted over stochastic simulations (as ribbons, or lines with low alpha etc.), as diagnostic tool to visualise the effect of stochasticity. Use the exact same transition rates so we can check if the mean of the stochastic simulations line up with the deterministic one.
• I noticed there are a few scripts that write function definitions using the same function name --- I'll tidy those up.
• design summary plots displaying simulated compartmental proportions, and the transition parameters + stochasticity, as a 'dashboard' to overview simulation outcomes w.r.t. these key input choices

[SeniorKate]

Saras and Kate contact details;

Github:

• smwindecker

• SeniorKate

Emails: kate.senior@thekids.org.au and saras.windecker@telethonkids.org.au

[varun-2901]

Varun – Weekly update

Done

• Moved my functions into R/ (simulate_sirs.R, sanity_check.R, plus utils)
• Added roxygen2 headers to all files in R/
• Standard 5-line headers in all scripts/ and auto-source via R.utils::sourceDirectory("R/")
• Fixed-β sanity check script (scripts/check_fixed_beta.R) — det vs mean stochastic
  RMSE I=0.00396; RMSE incidence=62.89 (pop=100k; β=0.18; γ=1/7; ω=1/30)
• Deterministic vs stochastic overlay plot function (R/plot_det_vs_stoch.R) + demo (scripts/demo_plot_det_vs_stoch.R)
• All figures routed to plots/ and ignored in git (.gitkeep added)
• Repo warnings fixed (ASCII hyphen in legend labels)

In progress / next

• PR from varun → main for review (references issues incl. Plot: deterministic vs stochastic overlay with quantile ribbon #7)
• Short README snippet on how to run: Rscript scripts/main.R, scripts/check_fixed_beta.R, scripts/demo_plot_det_vs_stoch.R

Notes

• Left Christy’s root .R files untouched to avoid conflicts; once she’s back, we can move those into R/ / scripts/ on her branch as per guidance.

[ChristyLAI99]

Work update:

1. Explored multi-population stochastic simulations. R/ simulate_sirs_multi_stoch.R plot_multi.R
2. Built plot_dashboard function.
3. Tidied functions into the R folder and added roxygen2-style docs.
4. Moved demos into scripts (standardized on demo_run.R)
5. Merged both branches into main under Kate and Saras’s guidance.

[ChristyLAI99]

Work update:

• Remove the contact matrix. check_contact.R and make_contact.R 9b23ca6
• Adjust the summarize_sim.R to cover the multiple populations. e6b5d5f
• Adjust the plotting style for Stochastic and Multiple population in plot_multi.R. 2464094
• Improved the plotting way of plot_dashboard.R, added self arrangement functions but not yet fix the dev vs stoch options inside. da7259d
• Added a draft Rmarkdown demo run for user, not yet add function form Varun. User-Demo.Rmd

What next?

• Improve the plot_dashboard.R like include the "dev vs stoch" plot for single population.
• REAME file
• Adding Unit test
• Finalise the User demo
• Tidy up all the functions documentations.