eks (version 1.0.7)

tidyst_kroc: Tidy and geospatial kernel receiver operating characteristic (ROC) curve

Description

Tidy and geospatial versions of kernel receiver operating characteristic (ROC) curve for 1- and 2-dimensional data.

Usage

tidy_kroc(data1, data2, ...)
st_kroc(x1, x2, ...)

Value

The output has the same structure as the 1-d kernel distribution estimate from *_kcde, except that fpr (\(x\)-variable) is the false positive rate (complement of specificity) and estimate is the true positive rate (sensitivity), rather than the usual estimation grid points and cdf values.

Arguments

data1, data2

data frames/tibbles of data values

x1,x2

sf objects with point geometry

...

other parameters in ks::kroc function

Details

A kernel ROC curve is a modification of the standard kernel distribution estimate where the two data samples are compared. For details of the computation and the bandwidth selection procedure of the kernel density ROC curve, see ?ks::kroc. The bandwidth matrix of smoothing parameters is computed as in ks::kcde per data sample.

Examples

## 2-d kernel ROC curve between unsuccessful and successful grafts
library(ggplot2)
data(hsct, package="ks")
hsct <- dplyr::as_tibble(hsct)
hsct <- dplyr::filter(hsct, PE.Ly65Mac1 >0 & APC.CD45.2>0)
hsct6 <- dplyr::filter(hsct, subject==6)   ## unsuccessful graft 
hsct6 <- dplyr::select(hsct6, PE.Ly65Mac1, APC.CD45.2)
hsct12 <- dplyr::filter(hsct, subject==12) ## successful graft 
hsct12 <- dplyr::select(hsct12, PE.Ly65Mac1, APC.CD45.2)
t1 <- tidy_kroc(data1=hsct6, data2=hsct12)
ggplot(t1, aes(x=fpr)) + geom_line(colour=1) 

## geospatial ROC curve between Grevillea species
data(wa)
data(grevilleasf)
hakeoides <- dplyr::filter(grevilleasf, species=="hakeoides")
paradoxa <- dplyr::filter(grevilleasf, species=="paradoxa")
s1 <- st_kroc(x1=hakeoides, x2=paradoxa)
ggplot(s1, aes(x=fpr)) + geom_line(colour=1)