Skip to contents

Love Plot for Covariate Balance Visualization

love_plot.Rd

Creates a publication-ready Love plot (standardized mean differences plot) comparing covariate balance before and after matching, weighting, or any other balancing procedure. Supports multiple adjustment stages.

Usage

love_plot(
  data_pre,
  data_post = NULL,
  treatment = NULL,
  weights_post = NULL,
  covariates = NULL,
  threshold = 0.1,
  abs_smd = TRUE,
  sort_by = c("pre", "post", "name"),
  title = "Covariate Balance: Love Plot",
  label_size = 3
)

Arguments

data_pre

Data frame or named numeric vector. Pre-adjustment covariate data (wide format) or pre-computed SMDs (named vector with covariate names as names).

data_post

Data frame or named numeric vector. Post-adjustment data or SMDs. If NULL (default), only pre-adjustment shown.

treatment

Character. Treatment variable name (only required when data_pre is a data frame).

weights_post

Numeric vector. Analytical weights for post-adjustment (if data_post is a data frame).

covariates

Character vector. Covariates to include. If NULL, all numeric covariates are used.

threshold

Numeric. The conventional balance threshold line (dashed). Default 0.1.

abs_smd

Logical. If TRUE (default), plot absolute SMD.

sort_by

Character. How to sort covariates: "pre" (default, by pre-SMD), "post", or "name".

title

Character. Plot title.

label_size

Numeric. Font size for covariate labels. Default 3.

Value

A ggplot2 Love plot.

References

Love, T. E. (2002). Displaying covariate balance after adjustment for selection bias: An application in healthcare. Annual Conference of the American Statistical Association.

Austin, P. C. (2009). Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in a propensity-score matched sample. Statistics in Medicine, 28(25), 3083–3107.

Examples

data(lalonde, package = "MatchIt")
m <- MatchIt::matchit(treat ~ age + educ + re74 + re75,
                      data = lalonde, method = "nearest")
love_plot(
  data_pre   = lalonde,
  data_post  = MatchIt::match.data(m),
  treatment  = "treat",
  covariates = c("age", "educ", "re74", "re75"),
  weights_post = MatchIt::match.data(m)$weights
)