psatControl: Creates a list of parameters for use with PSAT inference...
In ammeir2/PSAT: Inference after Selection with Aggregate Testing

Creates a list with additional parameters for use with mvnQuadratic, mvnLinear, and psatGLM.

psatControl(switchTune = NULL, nullMethod = c("RB", "zero-quantile"),
  nSamples = NULL, sgdStep = NULL, nsteps = 1000, trueHybrid = FALSE,
  rbIters = NULL, optimMethod = c("Nelder-Mead", "SGD"),
  truncPmethod = c("UMPU", "symmetric"), quadraticSampler = c("tmg",
  "PSAT"))

`switchTune`	tuning parameter for regime switching confidence intervals, should be a number between 0 and confidence_level.
`nullMethod`	method for compute global-null confidence intervals. Robins-Monroe (RB) should be used.
`nSamples`	number of samples to be taken from the null distribution.
`sgdStep`	number of stochastic gradient steps to take, only applicable for non-wald quadratic tests when `optimMethod` "SGD" is used.
`trueHybrid`	whether Robins-Monroe computation should be performed for computing confidence intervals for all confidence intervals or only when they may improve power.
`rbIters`	number of steps to take when computing confidence intervals with the Robins-Monroe procedure.
`optimMethod`	optimization method to be used when computing the conditional MLE in in inference after testing with a non-wald aggregate test. Nelder-Mead as implemented in the `optim` function.
`truncPmethod`	the type of test to use when computing the polyhedral p-values, options are either the UMPU test or a symmetric test.
`quadraticSampler`	which quadratic sampler to use? Choices are either the Hamiltionian Montel-Carlo method implemented in `rtmg`, or a Gibbs sampler implemented in this package.