hotelhd_control: Control Parameters for 'hotelhd'
In qpsy/hotelhd: High dimensional Hotelling's T2 test

Various parameters that control setting of tests. Currently, the maximum type tests (i.e., 'CLX', 'Z', 'M') only need control parameters. The followings are the parameters that should be considered for each method.

CLX: omegaGiven, omegaHat, omegaEst
M: omegaGiven, omegaHat, omegaEst, ndim, sub
Z: omegaGiven, omegaHat, omegaEst, B, Block

1
2
3

hotelhd_control(omegaGiven = NULL, omegaHat = c("omega", "identity"),
  omegaEst = c("clime"), ndim = 4L, sub = c("diag", "sub"), B = 500L,
  block = 1L)

`omegaGiven`	`NULL` or p by p matrix. When the default `NULL` is chosen, Ω matrix is estimated; see `omegaHat` and `omegaEst` parameters. Optionally, users can provide an estimate of Ω from an estimation method that this package does not offer.
`omegaHat`	character; 'omega' or 'identity'. A candidate estimate of Ω matrix for the maximum type methods. When the default of 'omega' is used, Ω is estimated. Users can choose estimation method by `omegaEst` parameter. Optionally, 'identity' can be choosen to set p by p Identity matrix as an estimate of Ω matrix. When the data is not high-dimensional, inverse of sample covariance is offered.
`omegaEst`	character; 'clime'. The estimation method for Ω matrix. Currently, only 'clime' is implemented; see `sugm`.
`ndim`	integer. The number of dimensions for generalized maximum type statistic, i.e., the method 'M'. The default is 4.
`sub`	character; 'diag' or 'sub'. Method about how to form `ndim` by `ndim` sub-matrix for the method 'M'. The default 'diag' uses diagonals of Ω to improve efficiency, and calculate results of 1 to `ndim`. The 'sub' extracts sub-matrix from Ω.
`B`	integer. The number of bootstrap statistics for 'Z'.
`block`	integer. A block size for blockwize multiplier bootstrap in the method 'Z'. The size should be smaller than min(n1, n2). The default value is 1.