caliperSize

The matching functions <code>fullmatch</code> and <code>pairmatch</code>
have a maximum problem size, based on the number of comparisons between treated
and control units. For a completely dense problem, in which every treated units
is compared to every control unit there are <code>length(treated) *
length(control)</code> comparisons. A caliper restricts which comparisons are valid,
disallowing matches of treated and control pairs that are too far apart. A
caliper can significantly decrease the size of a matching problem. The
<code>caliperSize</code> function reports exactly who many valid treated-control
comparisons remain after applying a caliper of the given width.

internal

Distance based bipartite matching using minimum cost flow, oriented
to matching of treatment and control groups in observational studies ('Hansen'
and 'Klopfer' 2006 <doi:10.1198/106186006X137047>). Routines are provided to
generate distances from generalised linear models (propensity score
matching), formulas giving variables on which to limit matched distances,
stratified or exact matching directives, or calipers, alone or in
combination.

Josh Errickson

optmatch

Functions for Optimal Matching

Ben Hansen

Mark Fredrickson

Josh Buckner

Adam Rauh

caliperSize function

<dl><dt>scores</dt>
<dd>A numeric vector of scores providing 1-D position of units</dd>
<dt>z</dt>
<dd>Treatment indicator vector</dd>
<dt>width</dt>
<dd>Width of caliper, must be positive</dd>
<dt>structure</dt>
<dd>Grouping factor to use in computation</dd></dl>

Arguments

(Internal) Determines how many other units fall within a caliper distance — caliperSize

<dl>

<dt>scores</dt>
<dd>A numeric vector of scores providing 1-D position of units</dd>


<dt>z</dt>
<dd>Treatment indicator vector</dd>


<dt>width</dt>
<dd>Width of caliper, must be positive</dd>


<dt>structure</dt>
<dd>Grouping factor to use in computation</dd>

</dl>

caliperSize: (Internal) Determines how many other units fall within a caliper distance

Description

Usage

Value

Arguments