GenCohortSingleOutcome: GenCohortSingleOutcome
In wlktan/surrogateSampling: Functions to conduct surrogate guided sampling designs

Generates a cohort with a single Bernoulli outcome with p features and length(betaZ) potential surrogates

1
2
3

GenCohortSingleOutcome(n = 1e+05, p = 250, p_nonzero = 30,
  beta0 = -14.75, betaZ = c(4.25, 2.25, 3.35), betaX, probZ = c(0.065,
  0.34, 0.62), probX, alpha0 = 0, alpha1 = 1, true.gamma = NULL)

`n`	cohort size
`p`	number of features
`p_nonzero`	number of nonzero features
`beta0`	intercept; controls prevalence of finding
`betaZ`	vector of true coefficients for surrogate Z
`betaX`	vector of true coefficients for features X
`probZ`	vector of probabilities of potential surrogate, i.e. P(Z=1)
`probX`	vector of probabilities of features, i.e. P(X=1)
`alpha0`	recalibration intercept
`alpha1`	recalibration slope
`true.gamma`	Additional features

A list with the following components: X: Matrix of dimension n x p for features Z: Matrix of dimension n x length(betaZ) for surrogates Y: Vector of 0/1 labels of length n trueAUC: AUC of the population if true probabilities were known true.beta: Vector of true coefficients

wlktan/surrogateSampling documentation built on May 23, 2019, 3:07 p.m.