tidyst_kcde {eks}

Tidy and geospatial kernel cumulative distribution and copula estimates

Description

Tidy and geospatial versions of kernel cumulative distribution estimates for 1- and 2-dimensional data, and kernel copula estimates for 2-dimensional data.

Usage

tidy_kcde(data, ...)
tidy_kcopula(data, ...)
st_kcde(x, ...)

Arguments

data

data frame/tibble of data values

x

sf object with point geometry

...

other parameters in ks::kcde, ks::kcopula functions

Details

For details of the computation of the kernel distribution and copula estimates, and of the bandwidth selector procedures, see ?ks::kcde, ?ks::kcopula.

Value

The outputs from *_kcde have the same structure as the kernel density estimate from *_kde, except that estimate indicates the cumulative distribution rather than the density values. Likewise for tidy_kcopula.

Examples

data(crabs, package="MASS")
## tidy 1-d distribution estimate per species
crabs1 <- dplyr::select(crabs, FL, sp)
crabs1 <- dplyr::group_by(crabs1, sp)
t1 <- tidy_kcde(crabs1)
gt1 <- ggplot2::ggplot(t1, ggplot2::aes(x=FL)) 
gt1 + ggplot2::geom_line(colour=1) + geom_rug_ks(colour=4) +
    ggplot2::facet_wrap(~sp)

## tidy 2-d copula estimate
crabs2 <- dplyr::select(crabs, FL, CW)
t2 <- tidy_kcopula(crabs2)
gt2 <- ggplot2::ggplot(t2, ggplot2::aes(x=FL, y=CW)) 
gt2 + geom_contour_filled_ks(colour=1, cont=seq(10,90,by=10))

## geospatial distribution estimate
data(wa)
data(grevilleasf)
paradoxa <- dplyr::filter(grevilleasf, species=="paradoxa")
paradoxa_crop <- sf::st_crop(paradoxa, xmin=4e5, xmax=8e5, ymin=6.4e6, ymax=6.65e6)
paradoxa_bbox <- sf::st_as_sfc(sf::st_bbox(paradoxa_crop))
xminb <- sf::st_bbox(paradoxa_crop)[1:2]
xmaxb <- sf::st_bbox(paradoxa_crop)[3:4]
s1 <- st_kcde(paradoxa_crop, xmin=xminb, xmax=xmaxb)

## base R filled contour plot
xlim <- c(1.2e5, 1.1e6); ylim <- c(6.1e6, 7.2e6)
plot(wa, xlim=xlim, ylim=ylim)
plot(paradoxa_bbox, add=TRUE, lty=3, lwd=2)
plot(s1, add=TRUE)

## geom_sf filled contour plot
gs1 <- ggplot2::ggplot(s1) + ggplot2::geom_sf(data=wa, fill=NA) + 
    ggthemes::theme_map() 
gs1 + ggplot2::geom_sf(data=paradoxa_bbox, linewidth=1.2, 
    linetype="dotted", fill=NA) +
    ggplot2::geom_sf(data=st_get_contour(s1), 
    ggplot2::aes(fill=label_percent(contlabel))) +
    ggplot2::scale_fill_viridis_d() +
    ggplot2::coord_sf(xlim=xlim, ylim=ylim)