optimize_hyperparameters: Optimize hyperparameters for GPRD
In duckmayr/gprd: Gaussian Process Regression for Regression Discontinuity Designs

Description Usage Arguments See Also

View source: R/optimize_hyperparameters.R

Optimize hyperparameters for GPRD

optimize_hyperparameters(
  x,
  y,
  cutoff = 0,
  estimator = "piecewise",
  b = NULL,
  B = NULL,
  method = "L-BFGS-B",
  control = list(type = 2),
  lower = 0.1,
  ...
)

`x`	A numeric vector; the forcing variable
`y`	A numeric vector; the outcome variable
`cutoff`	A numeric vector of length one; the treatment cutoff (default is 0)
`estimator`	A character vector of length one; should be one of "global" or "piecewise". The default is "global".
`b`	A numeric vector giving the prior mean for the coefficients for the mean function. If `NULL` (the default), all coefficients have a prior mean of 0.
`B`	A numeric matrix giving the prior covariance for the coefficients for the mean function. If `NULL` (the default), the prior covariance for the coefficients is a diagonal matrix, with the diagonal elements set equal to 10.
`method`	A character vector of length one giving the optimization routine to use; default is "L-BFGS-B", a box-constrained method. See `optim` for other options.
`control`	A list of control parameters for optim
`lower`	A numeric vector of length one giving the lower bound for the parameters. The default is 0.1.
`...`	Other arguments passed to `optim`