sfpcs_binary: This function computes the sfpc when the response variable is...
In YunlongNie/sFPCA: R package for supervised FPCA

Description Usage Arguments Examples

This function computes the sfpc when the response variable is binary.

1 2	sfpcs_binary(xmat, y, npc_select = 2, theta = 0.5, lambda = 10, timepts = NULL, xmat_new = NULL)

`xmat`	a matrix. Each column represents a sample.
`y`	a vector either 0 or 1. The binary response variable for each sample.
`npc_select`	an integer. The number of FPCs required.
`theta`	between 0 and 1. The weight parameter.
`lambda`	a positive number. The smoothing parameter.
`xmat_new`	a matrix. New dataset wish to predict on, each col corresponds to one sample.

## Not run: 
data(binary_dat);xmat = binary_dat$x;y=binary_dat$y
res = sfpcs_binary(xmat,y,npc_select=2,theta=1,xmat_new = xmat)
res$fitted
mean(y==res$fitted) # fitted accuracy
res$fitted # prediction on new data 
res$sfpcs # sFPCs 
res$beta_fd # coefficient function

## End(Not run)