Produce variables importance plot from an optint_by_group object. This plot has several features:
Estimates here are \(E(X | I=1) - E(X | I=0)\) and not cdf distances.
Star is added to variable name if there is a significant difference between at least two groups.
Estimates are standardized before they plotted (so different set of variables will have different standardization factor.)
# S3 method for optint_by_group
plot(x, plot.vars = "sig", graph.col = NULL,
alpha = 0.05, ...)an optint_by_group object.
which variables to plot? either a number (n) - indicating to plot the first n variables, "sig" (default) - plot only significant variables (here significant means that variabe is signifcant for all groups, or that there is significant heterogeneity), or a vector with names of variables to plot.
graph color/s.
significance level for the confidence intervals. also used in order to determine which variables are significant.
additional arguments.