quantile_envelope: Calculate quantile envelopes

In wgaul/wgutil: Willson's common ecological data analysis functions

Description

Calculate the pointwise quantile envelopes for comparison to observed data.

Usage

quantile_envelope(
  x,
  upper = 0.975,
  lower = 0.025,
  n.eval.points,
  type = "density"
)

Arguments

x	a dataframe with the observed values (column must be named "observed") and a column for each replicate to be used for quantile calculation
upper	a numeric value giving the desired upper quantile boundary (0 to 1)
lower	a numeric value giving the desired lower quantile boundary (0 to 1)
n.eval.points	integer giving the number of points at which to compute pointwise quantiles. This is passed to geom_density.
type	character vector indicating whether to produce envelope values as a density ("density") or expected value ("count")

Details

TODO: figure out difference b/t pointwise and complete envelopes

TODO: define a plot.env.envelope method

Value

a data frame with three columns holding the observed values, the upper envelope values, and the lower envelope values

See Also

geom_density

Examples

# make some example data
obs <- c(1:5, 7:8, 11:12, 19:25)
dat <- data.frame(observed = obs, s1 = sample(obs, replace = T),
                   s2 = sample(obs, replace = T),
                   s3 = sample(obs, replace = T),
                   s4 = sample(obs, replace = T),
                   s5 = sample(obs, replace = T))
# calculate the pointwise 95% quantile envelopes
result <- quantile_envelope(x = dat, n.eval.points = 20,
                            upper = 0.975, lower = 0.025)
plot(result$observed ~ result$x, type = "l", col = "black",
ylim = c(min(result$low_bnd), max(result$hi_bnd)),
main = "Quantile Envelope", ylab = "Density")
lines(result$low_bnd ~ result$x, col = "blue")
lines(result$hi_bnd ~ result$x, col = "blue")

wgaul/wgutil documentation built on June 1, 2020, 3:39 a.m.