bounds.logconc.internal: Computes upper identification interval under the assumption...

In jrzubizarreta/scbounds: Shape-constrained partial identification of a population mean under unknown probabilities of sample selection

Description

Computes upper identification interval under the assumption that F is log-concave.

Usage

bounds.logconc.internal(X, sampling.ratio = 5, xmin = NULL, xmax = NULL,
  buckets = 1000, alpha = 1/sqrt(length(X)))

Arguments

X	The observed data.
sampling.ratio	Bound on the sampling weights gamma.
xmin	Used to construct histogram representation.
xmax	Used to construct histogram representation.
buckets	Used to construct histogram representation.
alpha	Significance level used for KS bounds.

Value

mu.bound The upper bound for mu(x).

Xhat Unweighted empirical CDF of the data.

xvals Points at which Xhat is evaluated.

Xhat.upper KS bound for Xhat.

Lhat.upper Lhat function from paper.

Weighted version of Xhat that maximizes mu, subject to log-concavity

jrzubizarreta/scbounds documentation built on May 20, 2019, 5:26 p.m.