plot.matched.set: Plot the distribution of control unit weights

In PanelMatch: Matching Methods for Causal Inference with Time-Series Cross-Sectional Data

Plot the distribution of control unit weights

Description

The method creates a heatmap with the following characteristics. The heatmap grid is m x n, where m is the number of treated observations (as identified by i,t pairs) and n is the number of units. Treated observations represent the rows, and every unit in the data set form the columns. The figure then shows the calculated weights or distances (as specified) for each control unit within the matched set as identified by the row. Weights/distances that are missing or zero are not considered in the shading scheme and are both treated as NA for all practical purposes. Note that not all refinement methods will return a distance. Those that do also require verbose = TRUE in the PanelMatch specification. For example say (2, 5) is a treated observation and units 1, 4, 8 are matched as controls. Row i will represent (2,5) in the matrix, M. The columns indexed by w, x, and y, correspond to units 1, 4, and 8. M[i, w], M[i, x], M[i, y] then contain the weights or pairwise distances of units 1, 4, and 8 within that matched set.

Usage

## S3 method for class 'matched.set'
plot(
  x,
  ...,
  panel.data,
  type = "weights",
  include.missing = TRUE,
  low.color = "blue",
  mid.color = "white",
  high.color = "red",
  missing.color = "grey50"
)

Arguments

x	a matched.set object
...	Not used
panel.data	a PanelData object
type	character indicating whether or not weights or distances should be plotted
include.missing	logical. When TRUE, all units appear as columns, including those that are never included in any matched sets. When FALSE, only units that appear in at least one matched set are included.
low.color	option passed to ggplot2::scale_fill_gradientn(). The color representing the low weight/distance values.
mid.color	option passed to ggplot2::scale_fill_gradientn(). The color representing the medium weight/distance values.
high.color	option passed to ggplot2::scale_fill_gradientn(). The color representing the high weight/distance values.
missing.color	option passed to ggplot2::scale_fill_gradientn(). The color representing the missing/zero weight/distance values.

Value

returns a ggplot2::geom_tile() object producing a plot in alignment with the description above

Examples

 dem.panel <- PanelData(dem, "wbcode2", "year", "dem", "y")
 PM.results <- PanelMatch(panel.data = dem.panel, lag = 4,
                        refinement.method = "ps.match",
                        match.missing = TRUE,
                        covs.formula = ~ tradewb,
                        size.match = 5, qoi = "att",
                        lead = 0:4,
                        forbid.treatment.reversal = FALSE)

 mso <- extract(PM.results)
 plot(mso, panel.data = dem.panel)

PanelMatch documentation built on April 3, 2025, 6:34 p.m.