paper_simulations.R

#####
## Script for all paper simulations
#####

## 1) Simulations of raw data across a broad parameter array for two levels of complexity:
 ## -- no epidemiological controls, multiple epidemiological controls.
  ## -- i.e., a simple case to highlight findings uncomplicated by noise; a realistic case generating much noise
   ## (in the second case we consider mclust that generates multiple groups and how that varies from a forced two grouping)
    ## comparing the multiple grouping fit to the two gorup fit and how the multiple group fit corresponds to the two group fit

## 2) Method to summarize and choose a subset of these simulations that can then be funneled into three different pipelines
 
## 3) The three fitting pipelines to address the different conclusions we want to draw:
 ## A) Implications for choice of scale depending on varying skew (alongside distributional gap and proportion positive)
 ## B) What is gained and lost for individual and population level seropositivity by method
 ## C) What is gained and lost for epidemiological parameters by method