approx.cor {DecomposeR}

Correlation of time-series with different sampling rate

Description

Allows to correlate time-series having different sampling rate, if they have a comparable depth or time scale

Usage

approx.cor(xy1, dt1, xy2, dt2, plot = T, output = T, type = "p", ...)

Arguments

xy1

intensity values for the first data set

dt1

depth or time scale for the first data set

xy2

intensity values for the second data set

dt2

depth or time scale for the second data set

plot

whether to plot

output

whether to output

type

type of points in the plot (see help page of lines() for details)

...

additional parameters to feed to the lines() function

Value

a list of correlation ($cor), slope ($slope), intercept ($intercept) (two values for each: interpolation to fit dt1 and dt2 respectively), and of the xy1 and xy2 values, interpolated for dt1 ($df1) and df2 ($df2)

Examples

set.seed(42)

n <- 600
t <- seq_len(n)

p1 <- 30
p2 <- 240

xy.pure <- (1 + 0.6 * sin(t*2*pi/p2)) * sin(t*2*pi/p1)  + 2 * sin(t*2*pi/p2)

xy <- xy.pure + rnorm(n, sd = 0.5)

inter_dt <- round(runif(length(xy), min = 0.5, max = 1.5),1)

dt.pure <- cumsum(inter_dt)

keep <- runif(length(dt.pure)) < 0.5

xy <- xy[keep]
dt <- dt.pure[keep] + rnorm(sum(keep), -0.2, 0.2)

par(mfrow = c(1,2))

plot(xy, dt, type = "o", pch = 19)

plot(xy.pure, dt.pure, type = "o", pch = 19)

par(mfrow = c(1,1))

out <- approx.cor(xy, dt, xy.pure, dt.pure)

out$cor
out$slope
out$intercept