plot_residuals: Residuals plot

In pnolain/pmxploit: Pharmacometric tools for post-processing of mixed-effects models for continuous data

Residuals plot

Description

Plot residuals (e.g. CWRES, IWRES, NPDE) as either a scatterplot (versus an independent variable), an histogram (distribution) or a quantile-quantile plot.

Usage

plot_residuals(
  run,
  compartment = NULL,
  idv = "TIME",
  residuals,
  absolute_residuals = FALSE,
  keep_time_zero = FALSE,
  type = "scatterplot",
  smoothing_method = NULL,
  smoothing_se = TRUE,
  reference_value = NULL,
  histogram_bins = 30L,
  histogram_empirical_density = TRUE,
  histogram_reference_distribution = list(fun = dnorm, args = list(mean = 0, sd = 1)),
  qq_reference_distribution = list(fun = qnorm, args = list(mean = 0, sd = 1)),
  x_scale = "linear",
  y_scale = "linear",
  facet_scales = "free",
  transparency = FALSE,
  auto_legend = TRUE
)

Arguments

run	pmxploit NONMEM run object.
compartment	integer vector of the numbers of the compartments of residuals.
idv	character. Name of the column used as independent variable. Default is "TIME".
residuals	character vector of column names of residuals.
absolute_residuals	logical. If TRUE residuals are transformed to absolute values.
keep_time_zero	logical. If TRUE, does not ignore values at initiation (when TIME = 0). Default is FALSE.
type	character. Type of representation of the residuals. One of scatterplot, histogram and qq (Q-Q plot of a distribution of reference). Default is scatterplot.
smoothing_method	character. ggplot2 smoothing method.
smoothing_se	logical. Display confidence interval around smooth?
reference_value	numeric. Draws reference lines at +/- a particular value.
histogram_bins	integer. If type = "histogram", sets the number of bins. Default is 30.
histogram_empirical_density	logical. If TRUE, draws the empirical density over the histogram.
histogram_reference_distribution	logical. If TRUE, draws a theoretical density over the histogram.
x_scale	character. X-axis scale, one of "linear" or "log". Default is "linear".
y_scale	character. Y-axis scale, one of "linear" or "log". Default is "linear".
facet_scales	character. ggplot2 facet scales. Default is "free".
transparency	logical. Plot scatterplot dots with transparency, useful to avoid overplotting with large datasets. Default is FALSE.
auto_legend	logical. When TRUE (default), default pmxploit plotting options are applied (title, labels, axis...).

Value

A ggplot2 object.

Examples

EXAMPLERUN %>%
  plot_residuals(compartment = 2, idv = c("TIME", "PRED"), residuals = "CWRES")

EXAMPLERUN %>%
  group_by(CMT) %>%
  plot_residuals(compartment = 2:3, idv = c("TIME", "PRED"), residuals = "CWRES")

EXAMPLERUN %>%
  plot_residuals(compartment = 2, idv = c("TIME", "PRED"),
                 residuals = "CWRES", smoothing_method = "lm")

EXAMPLERUN %>%
  plot_residuals(compartment = 2, idv = c("TIME", "PRED"),
                 residuals = "CWRES", type = "histogram")
EXAMPLERUN %>%
  plot_residuals(compartment = 2, idv = c("TIME", "PRED"),
                 residuals = "CWRES", type = "qq")

EXAMPLERUN %>%
  group_by(STUD) %>%
  plot_residuals(compartment = 2, idv = c("TIME", "PRED"),
                 residuals = "CWRES", type = "qq")

pnolain/pmxploit documentation built on Jan. 31, 2024, 1:16 p.m.