BayesianReasoning 0.4.2

Major updates

• New plot_cutoff()
  • Shows healthy and sick distributions and shows FP, FN, TP and TN depending on a cutoff point
• New remove_layers_cutoff_plot() functions
  • Get's rid of layers of a cutoff_plot: FP, FN, TP or TN

Minor updates

• Fix for CRAN change in docType
• Use linewidth instead of size
• Faster implementation of a step in min_possible_prevalence

BayesianReasoning 0.4.0

Major updates

• Refactored the way we calculate PPV and NPV
• Added new parameter to control type of y scale (one_out_of) # 33
• Added new parameter to control x axis (limits_Sensitivity and limits_Specificity)
• Added multiple checks to catch corner conditions
• Renamed lots of parameters to make things more coherent

Minor updates

• Avoid acronyms
• Add bold label to area overlay
• Get rid of extra character in legend
• Translate all elements in area overlay
• Changed palette for NPV plots
• Add tests for new parameters and checks
• Add extra info in area overlay

Clean up, test, rinse and repeat

Clean up old comments, add more tests...

• 100% code coverage
• min_possible_prevalence() is now much more efficient
• Improvements to overlay = "line" en PPV_heatmap(), now using {ggforce} for labels
• Changed color palette for NPV
• Tweaked color palette for PPV

Testing, testing

CRAN release

It barely works

Initial release, includes 3 functions:

• PPV_heatmap.R: Plot PPV heatmaps
• PPV_diagnostic_vs_screening.R: Plot PPV for a diagnostic and a screening group
• min_possible_prevalence.R: Show minimum possible prevalence given the test characteristics

Right now it is a work in progress with lots of moving parts and horrible code.