fChange (version 0.2.1)

insert_change: Insert A Change In the Mean Function Of Functional Data

Description

This function inserts a change in the mean function to a given functional data sample. The change function can either be directly defined by the user or it can be generated based on the sum of first k basis functions defined by the fdobj. Once the change function is defined the change is inserted at the defined change location with a signal magnitude defined by signal to noise ratio, SNR. For more details on how these quantities are defined. See Aue, Rice and Sonmez (2017+).

Usage

insert_change(fdobj, change_fun = NULL, k = NULL, change_location, SNR,
  plot = TRUE, ...)

Arguments

fdobj

Functional data object of class 'fd'

change_fun

Self defined change function. It has to be a functional data object having the same number of basis functions.

k

Number of basis functions to be summed to construct the change function. It should be used when change_fun is not defined. It has to be less than number of basis functions.

change_location

Location of the change to be inserted. It is scaled to be in [0,1].

SNR

Signal to Noise Ratio to determine the magnitude of the change function that is being inserted.

plot

Plots the functional data before (blue) and after (red) the change.

...

Further information to pass

Value

fundata: functional data with an inserted change in the mean function

change_fun: inserted change function

plot: of the functional data with inserted change

Details

This function should only be used to artificially insert a change function to the mean of the functional data set either by defining a specific change function or generating the change function based from the basis functions.

References

Ramsay, James O., and Silverman, Bernard W. (2006), Functional Data Analysis, 2nd ed., Springer, New York.

Aue A., Rice G., Sonmez O. (2017+), Detecting and dating structural breaks in functional data without dimension reduction (https://arxiv.org/pdf/1511.04020.pdf)

See Also

Data2fd, fun_IID, fun_MA, fun_AR

Examples

# NOT RUN {
####################################
#first generate FAR(1) process
fdata = fun_AR(n=100, nbasis=25, Sigma=2^-(1:25))
# insert the change which is the sum of first 3 basis functions
# in the middle of the data with SNR=2
insert_change(fdata, k=3, change_location=0.5, SNR=2)

####################################
#first generate FAR(1) process
fdata = fun_AR(n=100, nbasis=25, Sigma=2^-(1:25))
# insert the change which is the 20th onservation
# in the middle of the data with SNR=2
insert_change(fdata, change_fun = fdata[20], change_location=0.5, SNR=2)
# }