lfqResample: Resampling of length-frequency data

In rschwamborn/fishboot: Bootstrap-based methods for the study of fish stocks and aquatic populations

Description

The function resamples the 'lfq' data by sample date. Sampling is done in a non-parametric manner that follows the relative frequencies of the original data, allowing for individual counts to be selected more than once (i.e. 'replace = TRUE' in sample), and resulting in total counts (by sample) equal to the original data.

Usage

lfqResample(lfq)

Arguments

lfq	a length frequency object of the class 'lfq'.

Value

a resampled version of the 'lfq' class dataset.

Examples

# load data
library(TropFishR)
data(alba)

# resample lfq data
alba_p <- lfqResample(alba)

# side-by-side plot
op <- par( mfcol = c(2,1), mar=c(4,4,2,1) )
plot(lfqRestructure(alba), Fname="rcounts")
mtext("original", side=3, line=0.25)
plot(lfqRestructure(alba_p), Fname="rcounts")
mtext("resampled", side=3, line=0.25)
par(op)

# relative difference
alba_diff <- alba
alba_diff$delta <- (alba$catch - alba_p$catch) / alba$catch
alba_diff$delta[is.na(alba_diff$delta)] <- 0
plot(alba, Fname = "catch", image.col = NA, hist.col=NA, draw = FALSE)
with(alba_diff, image(
  x=dates, y=midLengths, z=t(delta),
  zlim = max(abs(delta))*c(-1,1), col=rev(cm.colors(21)),
  add=TRUE
))
with(alba_diff, contour(x=dates, y=midLengths, z=t(delta), add=TRUE))
box()

rschwamborn/fishboot documentation built on July 27, 2019, 12:36 a.m.