varfunc: Estimate the variance function
In linulysses/mcfda: Mean and Covariance Estimation for Functional Data Analysis

Estimate the variance function

varfunc(
  Lt,
  Ly,
  newt = NULL,
  sig2 = NULL,
  method = c("PACE", "FOURIER"),
  mu = NULL,
  weig = NULL,
  ...
)

`Lt`	a list (for irregular design) or a vector (for regular design) of observations in time domain
`Ly`	a list (for irregular design) or a matrix (for regular design) of observations corresponding to `Lt`
`newt`	a set of locations where the variance function is estimated. If NULL, then set to `Lt`
`sig2`	variance of noise
`method`	estimation method, 'PACE' or 'FOURIER'
`mu`	an object generated by `meanfunc`.
`weig`	a vector of length(Lt). If NULL, then determined by the algorithm.
`...`	other parameters required depending on the `method` and `tuning`; see details

When method='PACE', additional parameters kernel and deg can be provided. bw as a scalar is optional. When bw is provided, the bandwidth is set to bw
When method='FOURIER', additional parameters q,rho,ext and domain are optional. If they are not provided, then they will be deduced from data or selected by the specified tuning method.