Elastic Partial Curve Matching

Find the optimal partial match of a template curve within a longer target curve using the elastic framework. This identifies the sub-interval of the target that best matches the template, along with the optimal warping.

Usage

elastic.partial.match(
  template,
  target,
  argvals.template,
  argvals.target,
  lambda = 0,
  min.span = 0.3
)

Arguments

template

Numeric vector of the template curve's values.

target

Numeric vector of the target curve's values.

argvals.template

Numeric vector of evaluation points for the template.

argvals.target

Numeric vector of evaluation points for the target.

lambda

Regularisation parameter controlling warping smoothness (default 0).

min.span

Minimum fraction of the target domain that the matched sub-interval must span (default 0.3).

Value

A list with components:

gamma

numeric vector of the warping function on the matched sub-interval

match_start

start of the matched sub-interval on the target domain

match_end

end of the matched sub-interval on the target domain

distance

elastic distance of the partial match

target_aligned

numeric vector of the matched and aligned target segment

References

Srivastava, A. and Klassen, E. (2016). Functional and Shape Data Analysis. Springer.

Examples

# \donttest{
t1 <- seq(0, 1, length.out = 50)
t2 <- seq(0, 2, length.out = 100)
template <- sin(2 * pi * t1)
target <- c(rep(0, 20), sin(2 * pi * seq(0, 1, length.out = 50)), rep(0, 30))
res <- elastic.partial.match(template, target, t1, t2)
# }