Plot density + ROC

Given positive and negative distributions, plot the corresponding densities and their ROC-curves.

Usage

plot_density_ROC(
  density_positive,
  cdf_positive,
  density_negative,
  quantile_negative,
  length.out = 500,
  xmin = -3,
  xmax = 3,
  colors_pos_neg_both = colors_pos_neg_both_default
)

Arguments

density_positive

function(x) {}: Returns the density given a specific value x. Refers to values coming from the "positive"-class.

cdf_positive

function(q) {...}: Returns the cumulative distribution value given a specific quantile q. Refers to values coming from the "positive"-class.

density_negative

function(x) {}: Returns the density given a specific value x. Refers to values coming from the "negative"-class.

quantile_negative

function(x) {}: Returns the density given a specific value x. Refers to values coming from the "negative"-class.

length.out

The number of simulated points for the plots from the distributions. length.out points will be generated between [xmin, xmax]. PURELY VISUAL, ROC curves are calculated based on the actual distributions!

xmin

Minimum value of simulated points for the plots from the distributions. length.out points will be generated between [xmin, xmax]. PURELY VISUAL, ROC curves are calculated based on the actual distributions!

xmax

Maximum value of simulated points for the plots from the distributions. length.out points will be generated between [xmin, xmax]. PURELY VISUAL, ROC curves are calculated based on the actual distributions!

colors_pos_neg_both

Colors for positive ("+"), negative ("-") and overlapping "+/-" areas of the densities

Value

patchwork'ed ggplots: left density, right ROC curve.

Examples

plot_density_ROC(
    density_positive = function(x) dnorm(x, mean = 0, sd = 1),
    cdf_positive = function(x) pnorm(x, mean = 0, sd = 1),
    density_negative = function(x) dnorm(x, mean = -1, sd = 1.5),
    quantile_negative = function(x) qnorm(x, mean = -1, sd = 1.5),
    xmin = -4, xmax = 7
)
#> $plot

#> 
#> $auc
#> 0.7104501 with absolute error < 8.1e-05
#>