GET (version 0.4)

create_curve_set: Create a curve_set object

Description

Create a curve_set object out of a list in the right form.

Usage

create_curve_set(curve_set, allfinite = FALSE, verbose = TRUE)

Value

An object of class curve_set containing the data. If the argument values are two-dimensional, then the curve_set is additionally a curve_set2d object.

Arguments

curve_set

A list containing the element obs, and optionally the elements r, sim_m and theo. See details.

allfinite

Logical. TRUE requires that all values of the curves must be finite (not infinite and not missing, see is.finite). FALSE allows for infinite or missing values in the curves. These infinite and missing values can have consequences for the subsequent analyses. A warning is given if infinite or missing values exists.

verbose

Logical flag indicating whether to warn about the content.

Details

The function is used to clump together the functional data in the form that can be handled by the other GET functions (forder, central_region, global_envelope_test etc.). The function create_curve_set takes care of checking the content of the data, and saves relevant information of the curves for global envelope methods to be used in particular for plotting the results with graphical interpretation.

obs must be either

• a vector containing the data function/vector, or

• a matrix containing the s data functions/vectors, in which case it is assumed that each column corresponds to a data function/vector.

If given, r describes the 1- or 2-dimensional argument values where the functions/vectors have been observed (or simulated). It must be either

• a vector,

• a data.frame with columns "x", "y", "width" and "height", where the width and height give the width and height of the pixels placed at x and y, or

• a data.frame with columns "xmin", "xmax", "ymin" and "ymax" giving the corner coordinates of the pixels where the data have been observed.

If obs is a vector, sim_m must be a matrix containing the simulated functions. Each column is assumed to correspond to a function, and the number of rows must match the length of obs. If obs is a matrix, sim_m is ignored.

If obs is a vector, theo can be given and it should then correspond to a theoretical function (e.g., under the null hypothesis). If present, its length must match the length of obs.

See Also

plot.curve_set, plot.curve_set2d

Examples

# 1d
cset <- create_curve_set(list(r = 1:10, obs = matrix(runif(10*5), ncol=5)))
plot(cset)
# 2d
cset <- create_curve_set(list(r = data.frame(x=c(rep(1:3, 3), 4), y=c(rep(1:3, each=3), 1),
                                           width=1, height=1),
                              obs = matrix(runif(10*5), ncol=5)))
plot(cset)