ham (version 1.2.0)

plot.assess: Prediction plot of treatment and control groups for DID and ITS models

Description

Provides partial prediction plots for treatment and control groups from difference-in-difference (DID) and interrupted time series (ITS) models. The graph will produce lines for treatment/intervention and control groups to gain understanding through a visual representation of the regression coefficients. By default, the treatment/intervention group is represented with a blue line, the control group is represented with a red line, and the counterfactual line, when available, is a dashed line. There are many options to change the plot.

Usage

# S3 method for assess
plot(
  x,
  y,
  xlim = NULL,
  ylim = NULL,
  xlab = NULL,
  ylab = NULL,
  main = NULL,
  lwd = NULL,
  col = NULL,
  tcol = NULL,
  cfact = FALSE,
  conf.int = FALSE,
  adj.alpha = NULL,
  add.means = FALSE,
  add.legend = NULL,
  legend = NULL,
  tgt = NULL,
  tgtcol = "gray",
  cex = 1,
  cex.axis = NULL,
  cex.lab = NULL,
  cex.main = NULL,
  cex.text = NULL,
  cex.legend = NULL,
  name = FALSE,
  coefs = FALSE,
  round.c = NULL,
  pos.text = NULL,
  arrow = FALSE,
  xshift = NULL,
  x.axis = NULL,
  ...
)

Value

plot of partial predictions for treatment and control groups.

Arguments

x

assess object. Either difference-in-difference or interrupted time series model with no covariate adjustment.

y

type of model, specify either 'DID' (difference-in-difference) or 'ITS' (interrupted time series). Will not accept other models.

xlim

specify plot's x-axis limits with a 2 value vector.

ylim

specify plot's y-axis limits with a 2 value vector.

xlab

a vector label for the x-axis.

ylab

a vector label for the y-axis.

main

the main title of the plot.

lwd

select the line width.

col

specify intervention and control group colors in a vector. Defaults to, if nothing selected, c("blue", "red") or "blue" for single-group Interrupted Time Series models.

tcol

specify treatment or interruption line color as a single character vector. Defaults to "gray" if nothing selected.

cfact

logical TRUE or FALSE that indicates whether a counterfactual line should be included. Defaults to FALSE.

conf.int

logical TRUE or FALSE that indicates whether a 95% confidence interval bands should be included. Defaults to FALSE.

adj.alpha

factor modifying the opacity alpha of the confidence interval bands, in the range of 0 to 1. Default is NULL; if conf.int=TRUE, defaults to 0.4.

add.means

adds group means by time period based on model data. Default is FALSE

add.legend

add a legend by selecting the location as "bottomright", "bottom", "bottomleft", "left", "topleft", "top", "topright", "right", "center". No legend if nothing selected.

legend

a character vector of length >= 1 to appear in the DID and ITS model legends to represent the Intervention and Control groups.

tgt

specify 1 or more values on the x-axis of where to add a target line when y='group'. Or 1 or more values on the y-axis of where to add a target line when y='time' or 'roll'. Default is NULL.

tgtcol

select a color for the target line. Default is 'gray'.

cex

A numerical value giving the amount by which plotting text and symbols should be magnified relative to the default of 1.

cex.axis

The magnification to be used for axis annotation relative to the current setting of cex.

cex.lab

The magnification to be used for x and y labels relative to the current setting of cex.

cex.main

The magnification to be used for main titles relative to the current setting of cex.

cex.text

The magnification to be used for the text added into the plot relative to the current setting of 1.

cex.legend

The magnification to be used for the legend added into the plot relative to the current setting of 1.

name

logical TRUE or FALSE that indicates whether coefficient names should be added to the plot. Default is FALSE. It is overridden if coefs = TRUE.

coefs

logical TRUE or FALSE that indicates whether coefficient names, values, and p-value significance symbols ('+' p<0.10; '' p<0.05; '' p<0.01; '' p<0.001) should be added to the plot. Default is FALSE. coefs = TRUE overrides name = FALSE.

round.c

an integer indicating the number of decimal places to be used for rounding coefficient values. Default is 2.

pos.text

a list of named integer value(s) between 1 to 4 indicating the position of the text added into the plot. List name(s) should use coefficient variable names.

arrow

logical TRUE or FALSE that indicates whether arrows and coefficient names should be added to visualize effects. Default is FALSE.

xshift

shifts one or two of some of the overlapping intervention associated arrows along the x-axis for a better view. Vector values of at least length 1 or 2 can be positive or negative. And xshift should be specified in the order of the coefficients. Only 1 or 2 of the furthest right, vertical lines for the intervention group is shifted (i.e., not left). One line is shifted when there is 1 treatment/interruption period and 2 shifts for 2 periods. (e.g., "DID" before "DID.Trend" for DID models with argument did="many").

x.axis

a vector of unique character or numeric values that makes up x-axis values to replace the intervention time variable values. This will be most helpful if you prefer current calendar years instead of values starting at 1 (e.g., x.axis= sort(unique(data$Year)) for 1900-1999, not 1-100). Must have equal lengths for unique x.axis values and unique replaced values. Default is NULL.

...

additional arguments.

Examples

am2 <- assess(formula= los ~ ., data=hosprog, intervention = "program",
topcode =NULL, int.time="month", regression="none", treatment= 5,
interrupt=c(5,9), did="two", its="two", newdata=TRUE, propensity=NULL)
plot(am2, "DID", add.legend="bottomleft", ylim=c(2, 8))  #DID model, basic plot
plot(am2, "ITS", add.legend="top", ylim=c(2, 8))         #ITS model, basic plot
plot(am2, "DID", add.legend="topleft", main="DID study", col=c("dodgerblue","magenta"),
ylim=c(2, 8), lwd=6, cex=3, cex.axis=2, cex.lab=1.5, cex.main=3, cex.text=2,
arrow=TRUE, xshift=0.02, coefs=TRUE, round.c=2 )
plot(am2, "ITS", add.legend="top", xlim=c(-.5, 13), ylim=c(2, 8), main="ITS study",
col=c("cyan","hotpink"), tcol="springgreen", lwd=7, cex=2, cex.axis=2, cex.lab=2,
cex.main=3, cex.text=1.2, cex.legend=1.25, name=FALSE, coefs=TRUE, round.c=1,
pos.text= list("txp5"=3, "post9"=4), arrow=TRUE, xshift=c(.5, 1.5),
cfact=TRUE, conf.int=TRUE, adj.alpha=0.2)
# ITS: USA's unemployment rate with a single group, 10 interruption model
key_time <- c(5, 14, 17, 29, 42, 59, 69, 73, 80,92)# key interruption periods
u110 <- assess(formula=rate ~ ., data=unemployment, intervention="usa",# model
               int.time="year", its="one", interrupt= key_time)
# Graph the results, note shift in intercept and slope during 1933's New Deal
plot(u110, "ITS", add.means = TRUE, coefs=TRUE, conf.int=TRUE, adj.alpha = .2,
     lwd=2, col="slategray", tcol= "orange", main="US unemployment rate",
     xlab="Years (1929-2024)", ylab="Proportion of labor market",
     cex.main=2, cex.axis = 1.5, cex.lab = 1.5, cex=2, cex.text = 1.25,
     pos.text=list("ITS.Time"=4, "post42"=1,"txp42"=3,"txp92"=3), x.axis=unemployment$Year)
for(i in 1:length(key_time)) {
  text(key_time[i], .22-(.01*i), cex=1.25, labels =
         paste0(unemployment[ key_time[i], "Year"], ": ", unemployment[ key_time[i], "event"]))
}