adaptive_knockoff: Adaptive knockoff
In zhimeir/adaptiveKnockoffs: Implementing knockoffs with side information.

This function takes W as an input and outpus the selection set generated by the adaptive knockoff procedure

adaptive_knockoff(
  W,
  U,
  fdr = 0.1,
  offset = 1,
  method = filter_gam,
  reveal_prop = 0.1,
  mute = TRUE
)

`W`	a p-dimensional vector of the feature importance statistics.
`U`	p-by-r vector of the side information.
`fdr`	The target FDR level(s). The default is 0.1.
`offset`	0 or 1 (default: 1).
`method`	The method used for evaluating the probability of being null. The default is random forest.