estimate_tables

This function estimates the total number of tables satisfying the margin constraints.
We recommend the readers to use this before determining the number of Monte Carlo simulations
for the sampling-based p-value calculation.

Implements exact, normally approximated, and sampling-based sensitivity
analysis for observational studies with contingency tables. Includes exact
(kernel-based), normal approximation, and sequential importance sampling (SIS)
methods using 'Rcpp' for computational efficiency. The methods build upon the
framework introduced in Rosenbaum (2002) <doi:10.1007/978-1-4757-3692-2> and
the generalized design sensitivity framework developed by Chiu (2025)
<doi:10.48550/arXiv.2507.17207>.

Elaine Chiu

sensitivityIxJ

Exact Nonparametric Sensitivity Analysis for I by J Contingency
Tables

estimate_tables function

<dl><dt>treatment.margins</dt>
<dd>A vector specifying the total number of subjects receiving each treatment level.
The first entry of this vector represents the number of subjects receiving the
first treatment level in the study.</dd>
<dt>outcome.margins</dt>
<dd>A vector specifying the total number of subjects showing each outcome level.
The first entry of this vector represents the number of subjects showing the
first outcome level in the study.</dd></dl>

Arguments

Estimate the Total Number of Tables — estimate_tables

<dl>

<dt>treatment.margins</dt>
<dd>A vector specifying the total number of subjects receiving each treatment level.
The first entry of this vector represents the number of subjects receiving the
first treatment level in the study.</dd>


<dt>outcome.margins</dt>
<dd>A vector specifying the total number of subjects showing each outcome level.
The first entry of this vector represents the number of subjects showing the
first outcome level in the study.</dd>

</dl>

estimate_tables: Estimate the Total Number of Tables

Description

Usage

Value

Arguments