permute_pointwiseT: Permutation test between two sets of curves

In snandi/Registration: Contains functions for registration algorithms, simulation and testing

Description

Permutation test between two sets of curves using T-type test statistic

Usage

permute_pointwiseT(Mat1, Mat2, Nperm = 200, argvals, q = 0.05, 
                   returnPlot = TRUE, TitleText = "")

Arguments

Mat1	Matrix 1 with each column being the discretized version of a separate curve
Mat2	Matrix 2 with each column being the discretized version of a separate curve
Nperm	Number of permutations. Default 200
argvals	The abscissa, the x-axis or the time-axis. The length of this should be the same as the number of rows of Mat1 and Mat2
q	0.05 or an appropriate α level
returnPlot	Whether to return the permutation distribution plot
TitleText	A text intended to be appended to ggtitle

Details

This is the permutation test, testing the hypothesis H_0: F_{Y_1} = F_{Y_2}

Value

pval	p-value of the permutation test
qval	q-value of the permutation test
Tobs	Observed value of the test statistic
Tvals	Observed pointwise values of T, along the argvals
Tnull	Null distribution of the test statistic
Tnullvals	Observed pointwise values of T, along the argvals, for all permutations
qvals.pts	Pointwise q-value, along the argvals
pvals.pts	Pointwise p-value, along the argvals
TPlot	Plot of Tobs in the Tnull distribution
Plot_pval	Plot of pointwise p-values, along the argvals

Note

This is an adaptation of the function fda::tperm.fd. The theory was first described in book mentioned in the references.

Author(s)

Subhrangshu Nandi; Statistics PhD student, UW Madison; [e-mail] or [e-mail]

References

Graves, S., Hooker, G., & Ramsay, J. (2009). Functional data analysis with R and MATLAB. (Sec 10.5.1)

See Also

fda::tperm.fd

Examples

data( growth, package = 'fda' )
Mat1 <- growth[['hgtm']]
Mat2 <- growth[['hgtf']]
Arguments <- growth[['age']]
PermTestResults <- permute_pointwiseT(
  Mat1 = Mat1, 
  Mat2 = Mat2, 
  Nperm = 200,
  argvals = Arguments,
  q = 0.05,
  returnPlot = TRUE,
  TitleText = 'Growths of boys and girls (Berkeley growth data)'
)
names(PermTestResults)

snandi/Registration documentation built on May 30, 2019, 5:04 a.m.