In metrumresearchgroup/pmplots: Plots for Pharmacometrics

Set up

library(pmplots)
library(dplyr)

data <- pmplots_data_obs()

A default plot

dv_pred(data)

Customize the axis label

dv_pred(data) + labs(x = "Predicted concentration (mg/L)")

Add Greek symbols to an axis label

There are a couple of ways to do this. The easiest is to have the latex2exp package installed and write the TeX right in the label.

dv_pred(data, x = "IPRED//Predicted concentration ($\\mu$g/mL)")

Otherwise, you can trigger R plotmath by putting !! at the start of the column label

dv_pred(data, x = "IPRED//!!'Predicted concentration (' * mu * 'g/mL)'")

Add a run number or other annotation to the plot

dv_pred(data) + labs(subtitle = "Run 1001")

Adopt the pmplots look and feel

At some point, you might have to create a plot outside of pmplots, but you still want to adopt the pmplots look and feel.

p <- ggplot(data = data, aes(PRED,DV)) + geom_point()
p

Use these helpers to style the plot

p + pm_theme() + pm_abline() + pm_smooth()

Customize the axis

If you start with this plot

dv_time(data)

you can use the xs and ys arguments to customize the x or y axis. Pass a list of named data that corresponds to arguments that you would pass to ggplot2::scale_x_continuous() or ggplot2::scale_y_continuous(). For example we can set the breaks like this

dv_time(data, xs = list(breaks = seq(0,240,30)))

But note there is an easier way using xby argument for any of the *_time() plots

dv_time(data, xby = 30)

You can put the y-axis on log scale

dv_time(data, ys = list(transform = "log10"))

Again note there is a log argument to this function that lets us do this more conveniently

dv_time(data, log = TRUE)

While we've had some convenient shortcuts built into some of the functions, this should illustrate how you can control anything related to the axis scales in your plot.

Specifically not that you should not do this

dv_time(data) + scale_y_continuous(transform = "log10")

This will result in a warning and clobber the scale that pmplots set. Always customize the axis through xs and ys.

Rotate axis labels

Use rot_x() or rot_y() and specify the angle to rotate the tick labels

cont_cat(data, x = "STUDYc", y = "NPDE") + rot_x(30)

or just set vertical = TRUE for rot_x()

cont_cat(data, x = "STUDYc", y = "NPDE") + 
  rot_x(vertical = TRUE)

metrumresearchgroup/pmplots documentation built on Oct. 15, 2024, noon