# antiExactMatch: Specify a matching problem where units in a common factor... In optmatch: Functions for Optimal Matching

 antiExactMatch R Documentation

## Specify a matching problem where units in a common factor cannot be matched.

### Description

This function builds a distance specification where treated units are infinitely far away from control units that share the same level of a given factor variable. This can be useful for ensuring that matched groups come from qualitatively different groups.

### Usage

``````antiExactMatch(x, z)
``````

### Arguments

 `x` A factor across which matches should be allowed. `z` A logical or binary vector the same length as `x` indicating treatment and control for each unit in the study. TRUE or 1 represents a treatment unit, FALSE of 0 represents a control unit. NA units are excluded.

### Details

The `exactMatch` function provides a way of specifying a matching problem where only units within a factor level may be matched. This function provides the reverse scenario: a matching problem in which only units across factor levels are permitted to match. Like `exactMatch`, the results of this function will most often be used as a `within` argument to `match_on` or another distance specification creation function to limit the scope of the final distance specification (i.e., disallowing any match between units with the same value on the factor variable `x`).

### Value

A distance specification that encodes the across factor level constraint.

`exactMatch`, `match_on`, `caliper`, `fullmatch`, `pairmatch`

### Examples

``````data(nuclearplants)

# force entries to be within the same factor:
em <- fullmatch(exactMatch(pr ~ pt, data = nuclearplants), data = nuclearplants)
table(nuclearplants\$pt, em)

# force treated and control units to have different values of `pt`:
z <- nuclearplants\$pr
names(z) <- rownames(nuclearplants)
aem <- fullmatch(antiExactMatch(nuclearplants\$pt, z), data = nuclearplants)
table(nuclearplants\$pt, aem)
``````

optmatch documentation built on Nov. 16, 2023, 5:06 p.m.