plot.expertsurv: Plot survival curves for the models fitted using 'fit.models'

In expertsurv: Incorporate Expert Opinion with Parametric Survival Models

Plot survival curves for the models fitted using fit.models

Description

Plot survival curves for the models fitted using fit.models

Usage

## S3 method for class 'expertsurv'
plot(...)

Arguments

...

Must include at least one result object saved as the call to the fit.models function. May include other optional parameters. These include: add.km: Whether the KM curve should be added. newdata: Specifies a profile of covariates (in the list newdata). Other possibilities are additional (mainly graphical) options: xlab: A string with the label for the x-axis (default = "time"). ylab: A string with the label for the y-axis (default = "Survival"). lab.profile: A (vector of) string(s) indicating the labels associated with the strata defining the different survival curves to plot. Defaults to the value used by the Kaplan Meier estimate given in fit.models. xlim: A vector determining the limits for the x-axis. colors: A vector of characters defining the colours in which to plot the different survival curves. lab.profile: A vector of characters defining the names of the models fitted. add.km: TRUE (whether to also add the Kaplan Meier estimates of the data). annotate: FALSE (whether to also add text to highlight the observed vs extrapolated data). legend.position: A vector of proportions to place the legend. Default to 'c(.75,.9)', which means 75% across the x-axis and 90% across the y-axis. legend.title: Suitable instructions to format the title of the legend; defaults to 'element_text(size=15,face="bold")' but other arguments can be added (using 'ggplot' facilities). legend.text: Suitable instructions to format the text of the legend; defaults to 'element_text(colour="black", size=14, face="plain")' but other arguments can be added (using 'ggplot' facilities). plot_opinion: TRUE will provide an illustration of the expert opinion at each time-point. plot_ci: Statistical uncertainty can be plotted using the plot_ci = TRUE argument and by specifying nsim equal to the number of desired simulations (for Bayesian models, this must be less than the total number of simulations from the posterior). By default, the confidence/credible intervals are plotted as dashed lines. If an area/ribbon plot is preferred, set ci_plot_ribbon = TRUE. nsim: Even if statistical uncertainty is not required in the plots, it is recommended that nsim is set to a reasonable number. If nsim = 1 by default, the maximum likelihood estimates or the posterior mean of the parameters will be used to plot the results. In most cases, this should suffice (particularly for maximum likelihood). However, the expected survival estimated by the full sampling distribution may be different from the estimate at its expectation/maximum likelihood estimate.

Value

A ggplot2 object of the survival curves.

Author(s)

Gianluca Baio

References

\insertRef

Baio.2020expertsurv

See Also

fit.models.expert

Examples

require("dplyr")
param_expert_example1 <- list()
param_expert_example1[[1]] <- data.frame(dist = c("norm","t"),
                                         wi = c(0.5,0.5), # Ensure Weights sum to 1
                                         param1 = c(0.1,0.12),
                                         param2 = c(0.15,0.5),
                                         param3 = c(NA,3))

timepoint_expert <- 14
data2 <- data %>% rename(status = censored) %>% mutate(time2 = ifelse(time > 10, 10, time),
                                                       status2 = ifelse(time> 10, 0, status))
example1_mle <- fit.models.expert(formula=Surv(time2,status2)~1,data=data2,
                                  distr=c("wph", "exp"),
                                  method="mle",
                                  pool_type = "log pool",
                                  opinion_type = "survival",
                                  times_expert = timepoint_expert,
                                  param_expert = param_expert_example1)


plot(example1_mle, add.km = TRUE, t = 0:30,plot_opinion = TRUE)

expertsurv documentation built on April 3, 2025, 10:37 p.m.