fun_icwt: Inverse of continuous wavelet transform

In WQM: Wavelet-Based Quantile Mapping for Postprocessing Numerical Weather Predictions

Inverse of continuous wavelet transform

Description

Inverse of continuous wavelet transform

Usage

fun_icwt(x.wave, dt, dj, flag.wav = "WaveletComp", scale = NULL)

Arguments

x.wave	input complex matrix.
dt	sampling resolution in the time domain.
dj	sampling resolution in the frequency domain.
flag.wav	String for two different CWT packages.
scale	Wavelet scales.

Value

reconstructed time series

References

fun_stoch_sim_wave in PRSim, Brunner and Furrer, 2020.

Examples

set.seed(100)

dt<-1
dj<-1/8
flag.wav <- switch(2, "wmtsa", "WaveletComp")

n <- 100
x <- rnorm(n)
x.wave <- t(WaveletComp::WaveletTransform(x=x)$Wave)
rec <- fun_icwt(x.wave, dt, dj, flag.wav)

x.wt <- WaveletComp::analyze.wavelet(data.frame(x=x),"x",dt=dt,dj=dj)
rec_orig <- WaveletComp::reconstruct(x.wt,only.sig = FALSE, plot.rec = FALSE)$series$x.r

### compare to original series
op <- par(mfrow = c(1, 1), mar=c(3,3,1,1), mgp=c(1, 0.5, 0))
plot(1:n, x, type="l", lwd=5, xlab=NA, ylab=NA)
lines(1:n, rec, col="red",lwd=3)
lines(1:n, rec_orig, col="blue", lwd=1)
legend("topright",legend=c("Raw","Inverse","Inverse_orig"),
       lwd=c(5,3,1),bg="transparent",bty = "n",
       col=c("black","red","blue"),horiz=TRUE)
par(op)

WQM documentation built on Oct. 11, 2024, 9:07 a.m.