fdars (version 0.3.3)

addError: Add Measurement Error to Functional Data

Description

Adds independent Gaussian noise to functional data observations.

Usage

addError(fdataobj, sd = 0.1, type = c("pointwise", "curve"), seed = NULL)

Value

An fdata object with added noise.

Arguments

fdataobj

An object of class fdata.

sd

Standard deviation of the Gaussian noise.

type

Type of noise:

pointwise

(Default) Independent noise at each evaluation point. Each f_i(t_j) gets independent noise.

curve

Common noise level per curve. Each curve gets a single random value added to all its points.

seed

Optional integer random seed for reproducibility.

See Also

simFunData, sparsify

Examples

t <- seq(0, 1, length.out = 100)
fd_clean <- simFunData(n = 20, argvals = t, M = 5, seed = 42)
fd_noisy <- addError(fd_clean, sd = 0.1)

oldpar <- par(mfrow = c(1, 2))
plot(fd_clean, main = "Clean Data")
plot(fd_noisy, main = "With Noise (sd = 0.1)")
par(oldpar)

# Higher noise level
fd_very_noisy <- addError(fd_clean, sd = 0.5)
plot(fd_very_noisy, main = "High Noise (sd = 0.5)")