forwardStop {selectiveInference}

ForwardStop rule for sequential p-values

Description

Computes the ForwardStop sequential stopping rule of G'Sell et al (2014)

Usage

forwardStop(pv, alpha=0.1)

Arguments

pv

Vector of **sequential** p-values, for example from fsInf or larInf

alpha

Desired type FDR level (between 0 and 1)

Details

Computes the ForwardStop sequential stopping rule of G'Sell et al (2014). Guarantees FDR control at the level alpha, for independent p-values.

Value

Step number for sequential stop.

Author(s)

Ryan Tibshirani, Rob Tibshirani, Jonathan Taylor, Joshua Loftus, Stephen Reid

References

Max Grazier G'Sell, Stefan Wager, Alexandra Chouldechova, and Rob Tibshirani (2014). Sequential selection procedures and Fflse Discovery Rate Control. arXiv:1309.5352. To appear in Journal of the Royal Statistical Society: Series B.

Examples

set.seed(33)
n = 50
p = 10
sigma = 1
x = matrix(rnorm(n*p),n,p)
beta = c(3,2,rep(0,p-2))
y = x%*%beta + sigma*rnorm(n)

# run forward stepwise
fsfit = fs(x,y)

# compute sequential p-values and confidence intervals
# (sigma estimated from full model)
out = fsInf(fsfit)
out

# estimate optimal stopping point
forwardStop(out$pv, alpha=.10)