plotConfidence: Plot confidence curve for (uE,E) set

View source: R/plotConfidence.R

plotConfidenceR Documentation

Plot confidence curve for (uE,E) set

Description

Plot confidence curve for (uE,E) set

Usage

plotConfidence(
  E,
  uE,
  stat = ErrViewLib::mue,
  normalize = TRUE,
  oracle = TRUE,
  probref = FALSE,
  conf_probref = FALSE,
  dist_probref = "Normal",
  rep_probref = 100,
  showUk = FALSE,
  plot = TRUE,
  score = TRUE,
  confStat = score,
  col = 2,
  type = c("l", "p"),
  add = FALSE,
  xlab = "k% discarded",
  xlim = NULL,
  ylab = ifelse(normalize, "MAE / MAE0", "MAE"),
  ylim = NULL,
  title = NULL,
  label = 0,
  showLegend = TRUE,
  legend = NULL,
  legLoc = "bottomleft",
  gPars = ErrViewLib::setgPars()
)

Arguments

E

(vector) prediction uncertainty, uE, or predicted value, V

uE

(vector) error or z-score

stat

(function) statistic to use

normalize

(logical) use normalized statistic

oracle

(logical) plot Oracle curve

probref

(logical) plot probabilistic reference (probref) curve

conf_probref

(logical) plot confidence band around probref curve

dist_probref

(string) model error distribution to generate Ideal curve

rep_probref

(integer) sampling repetitions for normal curve

showUk

(logical) plot secondary axis with u_k values; supersedes 'title'

plot

(function) produce plot ?

score

(logical) estimate DFPR and/or AUCO ?

confStat

(logical) estimate DFPR and/or AUCO ?

col

(integer) color index for the curve

type

(string) curve type: line ('l') or points ('p')

add

(logical) add confidence curve to previous plot

xlab

(string) x axis label

xlim

(vector) limits of the x axis

ylab

(string) y axis label

ylim

(vector) limits of the y axis

title

(string) a title to display above the plot

label

(integer) index of letter for subplot tag

showLegend

(logical) display legend

legend

(string) legend for the dataset

legLoc

(string) location of legend (see xy.coord)

gPars

(list) graphical parameters

Value

Plot confidence curve for (E,uE) set

Examples


  uE  = sqrt(rchisq(1000, df = 4))  # Re-scale uncertainty
  E   = rnorm(uE, mean=0, sd=uE)  # Generate errors
  plotConfidence(E,uE)



ppernot/ErrViewLib documentation built on June 1, 2024, 4:33 a.m.