plot_three_panel_two_folders: Plot three-panel figures for nested partially-latent model...
In zhenkewu/nplcm: Nested Partially-Latent Class Models (nplcm)

View source: R/plot-three-panel-two-folders.R

Visualize the model outputs for communicating how does data inform final etiology. Current code only works for single etiologies. Multiple etiology has not been coded, because we need special handling of pathogen ordering. Also, because multiple etiology has a different formula for model-based observed rate. DN: 1. current implementation: nplcm, BrS and SS. "Jfull" here is not the same as in other functions: it refers to the number of pathogens even if there are pathogens with only silver-standard data; in other functions, "Jfull" refers to the number of pathogens that have BrS data. 2. Missing data for BrS or SS are dropped when calculating observed measurement prevalences

1
2
3

plot_three_panel_two_folders(DIR_NPLCM1, DIR_NPLCM2,
  DIR_MEAS_PANEL = DIR_NPLCM2, index_for_order, SS_upperlimit = 1,
  eti_upperlimit = 1)

`DIR_NPLCM1`	File path 1 to the folder containing posterior samples
`DIR_NPLCM2`	File path 2 to the folder containing posterior samples
`DIR_MEAS_PANEL`	The file path that will appear measurement panels; Default is the second file path, `DIR_NPLCM2`.
`index_for_order`	The index of the list to be used for ordering the pathogens. Because each model output might give different ordering of pathogens (based on posterior mean).
`SS_upperlimit`	The upper limit of horizontal bar for the silver-standard subpanel (the middle panel). The default value is .25.
`eti_upperlimit`	The upper limit of horizontal bar for the etiology posterior subpanel (the rightmost panel). The default value is .4