plot.qgcompemmfit: Default plotting method for a qgcompemmfit object

In qgcompint: Quantile G-Computation Extensions for Effect Measure Modification

Default plotting method for a qgcompemmfit object

Description

Plot a quantile g-computation object from qgcompint. For qgcomp.emm.glm.noboot, this function will create a butterfly plot of weights. For qgcomp.emm.glm.boot and qgcomp.emm.glm.ee, this function will create a box plot with smoothed line overlaying that represents a non-parametric fit of a model to the expected outcomes in the population at each quantile of the joint exposures (e.g. '0' represents 'at the lowest quantile for every exposure')

Usage

## S3 method for class 'qgcompemmfit'
plot(
  x,
  emmval = NULL,
  suppressprint = FALSE,
  geom_only = FALSE,
  modelband = FALSE,
  flexfit = FALSE,
  modelfitline = FALSE,
  pointwisebars = TRUE,
  pointwiseref = 1,
  alpha = 0.05,
  ...
)

Arguments

x	"qgcompemmfit" object from qgcomp.emm.glm.noboot, qgcomp.emm.glm.boot, qgcomp.emm.glm.ee, or qgcomp.emm.cox.noboot functions
emmval	fixed value for effect measure modifier at which pointwise comparisons are calculated
suppressprint	If TRUE, suppresses the plot, rather than printing it by default (it can be saved as a ggplot2 object (or list of ggplot2 objects if x is from a zero- inflated model) and used programmatically)
geom_only	If TRUE, returns only the geometry (i.e. does not contain the entire plot object). Used for overlays. Only used for .ee and .boot methods. (default = FALSE)
modelband	(boot/ee only) If TRUE (FALSE=default), adds 95% prediction bands for E(Y|joint exposure) (the MSM fit)
flexfit	(boot/ee only) if TRUE (FALSE=default), adds flexible interpolation of predictions from underlying (conditional) model
modelfitline	(boot/ee only) If TRUE (FALSE=default), adds fitted (MSM) regression line of E(Y|joint exposure) to the smooth regression line plot
pointwisebars	(boot/ee only) If TRUE (TRUE=default), adds 95% error bars for pointwise comparisons of E(Y|joint exposure) to the smooth regression line plot
pointwiseref	(boot/ee only) integer (0=default): which category of exposure (from 1 to q) should serve as the referent category for pointwise comparisons? (default=1)
alpha	alpha level for all confidence intervals
...	Arguments (listed under “details” ) to underlying functions:

Details

The ... argument calls underlying plot functions, with arguments given here (similar to the plot function in the qgcomp package)

Value

If suppressprint=FALSE, then this function prints a plot specific to a "qgcompemmfit" object. If no bootstrapping is used, it will print a butterfly plot of the weights at the specified value of the modifier (set via emmval parameter) If bootstrapping is used, it will print a joint regression line for all exposures at the specified value of the modifier (set via emmval parameter)

If suppressprint=TRUE, then this function returns a "gg" (regression line) or "gtable" (butterfly plot) object (from ggplot2 package or gtable/grid packages), which can be used to print a ggplot figure and modify either of the above figures (see example below)

See Also

qgcomp.emm.glm.noboot, qgcomp.emm.glm.boot, qgcomp.emm.glm.ee, qgcomp.emm.cox.noboot

Examples

set.seed(50)
# linear model, binary modifier
dat <- data.frame(y=runif(50), x1=runif(50), x2=runif(50),
z=rbinom(50, 1, 0.5), r=rbinom(50, 1, 0.5))
(qfit <- qgcomp.emm.glm.noboot(f=y ~ z + x1 + x2, emmvar="z",
expnms = c('x1', 'x2'), data=dat, q=2, family=gaussian()))
plot(qfit, emmval = 1)
#
library(ggplot2)

# example with estimating equation approach
dat2 <- data.frame(y=runif(50), x1=runif(50), x2=runif(50),
z=sample(0:2, 50, replace=TRUE), r=rbinom(50, 1, 0.5))
dat2$z = as.factor(dat2$z)
(qfit4ee <- qgcomp.emm.glm.ee(f=y ~ z + x1 + x2, emmvar="z",
                          degree = 1,
                         expnms = c('x1', 'x2'), data=dat2, q=4, family=gaussian()))
pp0ee = plot(qfit4ee, emmval=0, suppressprint=TRUE)
pp1ee = plot(qfit4ee, emmval=1, suppressprint=TRUE)
pp2ee = plot(qfit4ee, emmval=2, suppressprint=TRUE)
pp1ee + theme_linedraw() # can use with other ggplot functions
# example with bootstrapping
dat2 <- data.frame(y=runif(50), x1=runif(50), x2=runif(50),
z=sample(0:2, 50, replace=TRUE), r=rbinom(50, 1, 0.5))
dat2$z = as.factor(dat2$z)
(qfit4 <- qgcomp.emm.glm.boot(f=y ~ z + x1 + x2, emmvar="z",
                          degree = 1, B = 20,
                         expnms = c('x1', 'x2'), data=dat2, q=4, family=gaussian()))
pp0 = plot(qfit4, emmval=0, suppressprint=TRUE)
pp1 = plot(qfit4, emmval=1, suppressprint=TRUE)
pp2 = plot(qfit4, emmval=2, suppressprint=TRUE)
pp1 + theme_linedraw() # can use with other ggplot functions

# overlay (fussy to work with)
#ppom <- ggplot_build(pp0 + pp1[2] + pp2[[2]] + scale_color_discrete(guide="none"))
ppom <- ggplot_build(pp0ee + pp1ee[2] + pp2ee[[2]] + scale_color_discrete(guide="none"))
ppom$data[[1]]$colour <- ppom$data[[2]]$colour <- "gray40" # emmval = 0 -> dark gray
ppom$data[[3]]$colour <- ppom$data[[4]]$colour <- "gray80" # emmval = 1 -> light gray
ppom$data[[5]]$colour <- ppom$data[[6]]$colour <- "black" # emmval = 2 -> black
xincrement = 0.025
ppom$data[[1]]$x <- ppom$data[[2]]$x <- ppom$data[[1]]$x - xincrement
ppom$data[[2]]$xmin <- ppom$data[[2]]$xmin - xincrement
ppom$data[[2]]$xmax <- ppom$data[[2]]$xmax - xincrement
ppom$data[[5]]$x <- ppom$data[[6]]$x <- ppom$data[[5]]$x + xincrement
ppom$data[[6]]$xmin <- ppom$data[[6]]$xmin + xincrement
ppom$data[[6]]$xmax <- ppom$data[[6]]$xmax + xincrement
plot(ggplot_gtable(ppom))

## Not run: 
library(gtable) # may need to separately install gtable
# example with no bootstrapping, adding object from bootstrapped fit
pp2 <- plot(qfit, emmval = 1, suppressprint=TRUE)
grid.draw(pp2)
# insert row on top that is 1/2 height of existing plot
pp2b = gtable::gtable_add_rows(pp2, heights=unit(0.5, 'null') , pos = 0)
# add plot to that row
pp3 = gtable::gtable_add_grob(pp2b, ggplot2::ggplotGrob(pp), t=1, l=1, r=2)
grid.draw(pp3)

## End(Not run)

qgcompint documentation built on April 3, 2025, 7:49 p.m.