General plot for the envelope, instantaneous frequency (period) and identity tuning of an intrinsic mode function (IMF)
plot_imf(
pulse,
dtlim = NULL,
xylim = NULL,
flim = NULL,
fclim = NULL,
dtline = NULL,
fline = NULL,
fcline = NULL,
vertical = FALSE,
n = 10,
at.maj = NULL,
ls = list(type = "o", pch = 19),
le1 = list(lwd = 2),
le2 = list(lty = 2),
lid = list(type = "p", pch = 19),
lcos = list(),
ldt = list(lty = 5, lwd = 2),
lf = list(lty = 5),
lfc = list(lty = 5),
box = TRUE
)a pulse object
the boundaries for the plots, respectively for the depth/time, amplitude, frequency and freqeuncy carrier
coordinates to add vertical/horizontal lines
whether to have the depth/time [dt] axis vertically
the the number of intervals defined by minor ticks (geologist convention) or horizontaly (climatologist convention)
the positions at which major tick-marks are to be drawn.
lists of parameters to feed lines, for the original signal, the upper and lower envelope, the identity tuning, and the cosine line in the identity tuning
lists of parameters to provide the abline function (makes personalised lines for you to have a better grasp of the data).
whether to draw boxes around the plots
the line in the identity tuning plot is a genuine cosine, independent from the signal. This is evident when riding waves generate dephasing.
n <- 600
t <- seq_len(n)
p1 <- 30
p2 <- 40 * 21
am <- sin(t*2*pi/p2 + 50) + 0.03
xy <- sin(t*2*pi/p1 + 50) * 3 * am
int <- c(rep(1, 99 + 100), seq(1,3,2/100), seq(3,1,-2/100), rep(1,100 + 99))
dt <- cumsum(int)
samp <- approx(dt, xy, xout = seq(1,802, by = 2))
xy <- samp$y
dt <- samp$x
e <- normalise(m = xy, dt = dt)$a
cond <- dt < 75
xy <- xy[!cond]
dt <- (dt[!cond] - 75) / 1.2
e <- e[!cond]
dq <- dq.algorithm(xy/e, dt)
pulse <- as.pulse(dt = dt, m = xy, f = dq$f, a = e, idt = dq$idt,
repl = 1)
plot_imf(pulse, fline = 25, dtline = c(222, 489))
Run the code above in your browser using DataLab