This is a Summer Research collaboration with professor Nathaniel Osgood's lab on Public Health and Epidimology, University of Saskachewan. link:https://drive.google.com/drive/folders/12xhx-ZhSkbtHluF-6Dgw2TmBYE8l3X1S?usp=sharing
| What Olivia did | |
|---|---|
| AGEGROUP Added AgeGroup structure to the original model | 1.child(0-12), teen(12-18), old(18+)2.From 14 stocks to 38 stocks3.From 2 dimension (gender) to 6 dimension (gender*age)4.Modified remandModel.c, remandModel.h files5.Added "ageout_child" and "ageout_teen" flow and modified the integration from 1 section to 3 section (3 forloops). |
| Updated likelihood function to log likelihood function | 1.Solved the "newWeight = nan" problem, it was because one of the likelihood = nan. |
| Updated graphs in R | 1.Changed the graphs from female, male to female_child, female_teen, female_old, male_child, male_teen, male_old2.Modified RInterfaceRemand file3.Note: Indices are off in the R file: the index in R should be index in remandModel.h+1 |
| VARY | 1.arrest rate varies between a minimum (0.015) and a maximum value (0.03) |
| Future work | |
|---|---|
| Add new empirical data | Column1:year (for double check)Column2:inPolice teen (detention)Column3:inPolice old (detention) Remand and sentenced custody for child = 0 |
| Double check "countTimePoints"<- 133 | I changed countTimePoints from 299 to 133 because: stopifnot(length(arrayTimesForTimepoints) * countObservables == length(arrayObservedValuesYByTimepointAndObservable)) |
| Adjust initial value for stocks | Initial detention value |
| How to run PMCMC | Run PMCMC: In terminal: Set the path to be in the foldercd + path docker-compose up -d docker exec -it rstudio bash cd code/ ulimit -s unlimited Rscript RInterfaceRemand.R |
| How to fix PMCMC | ls -l docker ps docker container rm -f rstudio docker exec -it rstudio bash If modified the C file: make cleanmake |
| PF | |
|---|---|
| IZ_V15 | 1.Has measles, chicken pox and influenza with respective empirical values.2.Immunity is fixed 3.Added Calibration (Optimization Experiment)4.I ignored the histograms and plots so that calibration can run faster |
| Rubella V1 | 1.Has measles, chicken pox and Rubella with respective empirical values.2.Has calibration and histograms |
| Rubella V2 | 1.Has calibration and but no histograms |