ccor: Curve correlation

In ctmva: Continuous-Time Multivariate Analysis

Curve correlation

Description

Inputs raw data representing two curves, applies penalized B-spline smoothing to the two curves, and computes the curve correlation between them via a call to cor.ct.

Usage

ccor(y, time, curve = NULL, k = 15, min.overlap = 0, min.n = 8)

Arguments

y	either a vector or a two-column matrix, without missing values; see Details
time	a vector of time points
curve	curve indicator; see Details
k	number of B-spline basis functions
min.overlap	minimum overlap of the two curves' time ranges
min.n	minimum number of observations per curve

Details

If y is a two-column matrix, the two curves are observed at the time points given by time; in this case length(time) must equal nrow(y), and curve is ignored. If y is a vector, it must have the same length as both time and curve. In this case y contains the observations on both curves, while elements of time and curve identify the observation time and the curve being observed, respectively.

Value

A list with components

y, time	the supplied y and time
mod1, mod2	models for the two curves, outputted by gam
fd1, fd2	functional data objects (see fd) for the two curves
estcor	estimated curve correlation

Author(s)

Philip Tzvi Reiss <[e-mail]>, Noemi Foa, Dror Arbiv and Biplab Paul <[e-mail]>

See Also

cor.ct, b.spline

Examples

# see example for ccor_posim

ctmva documentation built on Nov. 20, 2025, 5:06 p.m.