nonrandom (version 1.42)

plot.stdf: Graphical check for standardized differences of covariates

Description

Plot standardized differences of covariates

Usage

## S3 method for class 'stdf':
plot(x, sel = NULL, plot.alpha = TRUE,
  mymar = c(5,8,4,2), pch.p = c(1,5), col.p = c("black", "red"),
  colorspace = NULL, cex.p = 1.25, line.stdf = 1, line.alpha = 4,
  with.legend = TRUE, legend.label = c("before", "after"),
  legend.cex = 1, legend.xy = NULL, ...)

Arguments

x

an object of class 'bal.matched.data.frame', 'bal.matched.data.frames' or 'bal.matched.pscore'. The previous use of ps.balance(..., method='stand.diff', ...) is needed.

sel

a vector of strings indicating covariates to be checked. The default is NULL, i.e. all variables selected previously in ps.balance() are plotted.

plot.alpha

a logical value indicating whether a vertical line for the significance level chosen in ps.balance() should be plotted.

mymar

a graphical parameter, see par().

pch.p

a vector of two integers indicating the symbols.

col.p

a vector of two strings indicating symbol colors.

colorspace

a logical value indicating whether rainbow_hcl(20) (if TRUE) or grey.colors(20) (if FALSE) should be used (argument '20' indicate that 20 colors are pre-specified and can be selected using col.p). The default i

cex.p

a numeric indicating the size of symbols.

line.stdf

an integer indicating the line type connecting the sympols.

line.alpha

an integer indicating the line type presenting the significance level.

with.legend

a logical value indicating whether the legend is given.

legend.label

a vector of two strings indicating the labels.

legend.cex

a numeric indicating the size of the legend font.

legend.xy

a vector of two integers indicating x- and y-coordinates for legend position.

...

further arguments for graphics.

Details

Standardized difference are proposed to check the balance of covariates after matching. The function plot.stdf() offers a graphical check of it presenting standardized differences of covariates between treatment groups before and after matching. The usage of plot.stdf() requires the previous use of ps.balance(..., method='stand.diff', ...) and is only available when matching is done before via ps.match().

See Also

plot

Examples

## STU1
data(stu1)
stu1.ps <- pscore(data    = stu1, 
                  formula = therapie~tgr+age)

stu1.match <- ps.match(object          = stu1.ps,
                       ratio           = 2,
                       caliper         = 0.5,
                       givenTmatchingC = FALSE,
                       matched.by      = "pscore",
                       setseed         = 38902)

stu1.bal.match <- ps.balance(object = stu1.match,
                             sel    = c("tgr","age"),
                             method = "stand.diff",
                             alpha  = 20)

plot.stdf(x          = stu1.bal.match,
          main       = "Stu1 study: Standardized differences",
          cex.axis   = 1.2,
          legend.cex = 1.3,
          cex.main   = 1.5,
          las        = 1,
          col.p      = c("black", "gray"))