ggsurv: Survival curves with ggplot2

In osofr/estimtr: Streamlined Estimation for Static, Dynamic and Stochastic Treatment Regimes in Longitudinal Data

Description

This function produces plots of the survival point estimates using ggplot2. As a first argument it needs a list of estimates created by the one of the estimation function of the stremr package. See http://ggobi.github.io/ggally/#ggallyggsurv for additional modifications of the resulting output plot.

Usage

ggsurv(
  estimates,
  CI = TRUE,
  CI_line = FALSE,
  plot_cens = TRUE,
  surv_col = "gg.def",
  cens_col = "gg.def",
  lty_est = 1,
  shape_est = seq_along(estimates),
  lty_ci = 2,
  size_est = 0.5,
  size_ci = size_est,
  size_pt = size_est + 0.3,
  cens_size = 2,
  cens_shape = 3,
  back_white = TRUE,
  xlab = "Time",
  ylab = "Survival",
  main = "",
  legend_pos = "right",
  surv_name = "St." %+% estimates[[1L]][["est_name"]][1L],
  SE_name = "SE." %+% estimates[[1L]][["est_name"]][1L],
  order_legend = TRUE,
  t_int_sel = NULL,
  ymin = NULL,
  ...
)

Arguments

estimates	A list, one item per regime / intervention. Each list item must be a data.frame containing the survival estimates by time for a single regime / intervention.
CI	should a 95% confidence interval be plotted? Defaults to TRUE. Uses the standard error estimates provided as a separate column of the input data.
CI_line	When TRUE the 95% CIs will be plotted as a line function (same as main plot type). When FALSE the 95% CIs are plotted using ggplot2::geom_ribbon.
plot_cens	mark the censored observations?
surv_col	colour of the survival estimate. Defaults to black for one stratum, and to the default ggplot2 colours for multiple strata. Length of vector with colour names should be either 1 or equal to the number of strata.
cens_col	colour of the points that mark censored observations.
lty_est	linetype of the survival curve(s). Vector length should be either 1 or equal to the number of strata.
shape_est	shape type of the survival point estimates. Vector length should be either 1 or equal to the number of strata.
lty_ci	linetype of the bounds that mark the 95% CI.
size_est	line width of the survival curve
size_ci	line width of the 95% CI
size_pt	point size of the survival estimate at each time-point
cens_size	point size of the censoring points
cens_shape	shape of the points that mark censored observations.
back_white	if TRUE the background will not be the default grey of ggplot2, but will be white with borders around the plot. Defaults to TRUE
xlab	the label of the x-axis.
ylab	the label of the y-axis.
main	the plot label.
legend_pos	Either the coordinates of the legend position inside the plot (e.g., (0.9, 0.2)) or the character word denoting the legend orientation with respect to the plot (e.g., "bottom", "right" or "left").
surv_name	The name of the column containing the survival estimates.
SE_name	The name of the column containing the standard errors (SE) for each time-point estimate of survival.
order_legend	Set to TRUE to order the legend display by final survival time (highest first).
t_int_sel	The subset of time-point indices for which survival should be plotted.
ymin	The minimum value of the y axis. The default (ymin=NULL) is to use ggplot to automatically adjust the limits of the y-axis.
...	Additional arguments (not used).

Value

An object of class ggplot

Author(s)

Original R code by Edwin Thoen [e-mail], modified by Oleg Sofrygin [e-mail]

osofr/estimtr documentation built on Jan. 25, 2022, 8:05 a.m.