pairsplot {designmatch} | R Documentation |
Pairs plot for visualizing matched pairs
Description
Function for visualizing matched pairs in two dimensions.
Usage
pairsplot(cov1, cov2, t_id, c_id, xlab, ylab, main)
Arguments
cov1 |
a vector for the covariate to be plotted on the x axis. |
cov2 |
a vector for the covariate to be plotted on the y axis. |
t_id |
a vector of indexes of the treated units. |
c_id |
a vector of indexes of the matched controls. |
xlab |
a string specifying the label of the x axis. |
ylab |
a string specifying the label of the y axis. |
main |
a string specifying the main title of the plot. |
Details
pairsplot
is a function for visualizing matched pairs in two dimensions, usually defined by two of the matching covariates.
Matched pairs are connected by line segments.
Horizontal and vertical lines show the means of treated units and matched controls for each of the covariates.
Among others, pairsplot
can be useful for visualizing near/far matches, e.g. when building a stronger instrumental variable (Baiocchi et al., 2010).
Author(s)
Jose R. Zubizarreta <zubizarreta@hcp.med.harvard.edu>, Cinar Kilcioglu <ckilcioglu16@gsb.columbia.edu>.
References
Baiocchi, M., Small, D., Lorch, S. and Rosenbaum, P. R. (2010), "Building a Stronger Instrument in an Observational Study of Perinatal Care for Premature Infants," Journal of the American Statistical Association, 105, 1285-1296.
Examples
# Load data
data(germancities)
# Sort and attach data
germancities = germancities[order(germancities$treat, decreasing = TRUE), ]
attach(germancities)
# Treatment indicator
t_ind = treat
# Indexes of the treated units
t_id = which(t_ind == 1)
# Indices of the matched controls (obtained using bmatch in designmatch)
c_id = c(67, 75, 39, 104, 38, 93, 79, 59, 64, 55, 106, 99, 97, 61, 82,
57, 76, 47, 46, 49)
# pairsplot
pairsplot(rubble, flats, t_id, c_id, "Rubble", "Flats", "")