Plot Method for survfit.JMbayes and survfit.mvJMbayes Objects
Produces plots of conditional probabilities of survival.
## S3 method for class 'survfit.JMbayes' plot(x, estimator = c("both", "mean", "median"), which = NULL, fun = NULL, invlink = NULL, conf.int = FALSE, fill.area = FALSE, col.area = "grey", col.abline = "black", col.points = "black", add.last.time.axis.tick = FALSE, include.y = FALSE, main = NULL, xlab = NULL, ylab = NULL, ylab2 = NULL, lty = NULL, col = NULL, lwd = NULL, pch = NULL, ask = NULL, legend = FALSE, ..., cex.axis.z = 1, cex.lab.z = 1, xlim = NULL) ## S3 method for class 'survfit.mvJMbayes' plot(x, split = c(1, 1), which_subjects = NULL, which_outcomes = NULL, surv_in_all = TRUE, include.y = TRUE, fun = NULL, abline = NULL, main = NULL, xlab = "Time", ylab = NULL, zlab = "Event-Free Probability", include_CI = TRUE, fill_area_CI = TRUE, col_points = "black", pch_points = 1, col_lines = "red", col_lines_CI = "black", col_fill_CI = "lightgrey", lwd_lines = 2, lty_lines_CI = 2, cex_xlab = 1, cex_ylab = 1, cex_zlab = 1, cex_main = 1, cex_axis = 1, ...)
x: an object inheriting from class survfit.JMbayes or class survfit.mvJMbayes.estimator: character string specifying, whether to include in the plot the mean of the conditional probabilities of survival, the median or both. The mean and median are taken as estimates of these conditional probabilities over the M replications of the Monte Carlo scheme described in survfitJM.which: an integer or character vector specifying for which subjects to produce the plot. If a character vector, then is should contain a subset of the values of the idVar variable of the newdata argument of survfitJM.which_subjects: an integer vector specifying for which subjects to produce the plot.split: a integer vector of length 2 indicating in how many panels to construct, i.e., number of rows and number of columns.which_outcomes: integer vector indicating which longitudinal outcomes to include in the plot.surv_in_all: logical; should the survival function be included in all panels.fun: a vectorized function defining a transformation of the survival curve. For example, with fun=log the log-survival curve is drawn.abline: a list with arguments to abline().invlink: a function to transform the fitted values of the longitudinal outcome.conf.int, include_CI: logical; if TRUE, then a pointwise confidence interval is included in the plot.fill.area, fill_area_CI: logical; if TRUE the area defined by the confidence interval of the survival function is put in color.col.area, col_fill_CI: the color of the area defined by the confidence interval of the survival function.col.abline, col.points, col_points, col_lines, col_lines_CI: the color for the vertical line and the points when include.y is TRUE.add.last.time.axis.tick: logical; if TRUE, a tick is added in the x-axis for the last available time point for which a longitudinal measurement was available.include.y: logical; if TRUE, two plots are produced per subject, i.e., the plot of conditional probabilities of survival and a scatterplot of his longitudinal measurements.main: a character string specifying the title in the plot.xlab: a character string specifying the x-axis label in the plot.ylab: a character string specifying the y-axis label in the plot.ylab2: a character string specifying the y-axis label in the plot, when include.y = TRUE.zlab: a character string specifying the z-axis (vertical right-hand side) label in the plot.lty, lty_lines_CI: what types of lines to use.col: which colors to use.lwd, lwd_lines: the thickness of the lines.pch, pch_points: the type of points to use.ask: logical; if TRUE, the user is asked before each plot, see par().legend: logical; if TRUE, a legend is included in the plot.cex.axis.z, cex.lab.z: the par cex argument for the axis at side 4, when include.y = TRUE.cex_xlab, cex_ylab, cex_zlab, cex_main, cex_axis: the par cex argument for the axis in side 1 (x-axis), side 2 (y-axis), side 4 (z-axis) and the title of the plot.xlim: the par xlim argument....: extra graphical parameters passed to plot().Rizopoulos, D. (2016). The R package JMbayes for fitting joint models for longitudinal and time-to-event data using MCMC. Journal of Statistical Software 72(7) , 1--45. doi:10.18637/jss.v072.i07.
Rizopoulos, D. (2012) Joint Models for Longitudinal and Time-to-Event Data: with Applications in R. Boca Raton: Chapman and Hall/CRC.
Rizopoulos, D. (2011). Dynamic predictions and prospective accuracy in joint models for longitudinal and time-to-event data. Biometrics 67 , 819--829.
Dimitris Rizopoulos d.rizopoulos@erasmusmc.nl
survfitJM
## Not run: # we construct the composite event indicator (transplantation or death) pbc2$status2 <- as.numeric(pbc2$status != "alive") pbc2.id$status2 <- as.numeric(pbc2.id$status != "alive") # we fit the joint model using splines for the subject-specific # longitudinal trajectories and a spline-approximated baseline # risk function lmeFit <- lme(log(serBilir) ~ ns(year, 2), data = pbc2, random = ~ ns(year, 2) | id) survFit <- coxph(Surv(years, status2) ~ drug, data = pbc2.id, x = TRUE) jointFit <- jointModelBayes(lmeFit, survFit, timeVar = "year") # we will compute survival probabilities for Subject 2 in a dynamic manner, # i.e., after each longitudinal measurement is recorded ND <- pbc2[pbc2$id == 2, ] # the data of Subject 2 survPreds <- vector("list", nrow(ND)) for (i in 1:nrow(ND)) { survPreds[[i]] <- survfitJM(jointFit, newdata = ND[1:i, ]) } # the default call to the plot method using the first three # longitudinal measurements plot(survPreds[[3]]) # we produce the corresponding plot par(mfrow = c(2, 2), oma = c(0, 2, 0, 2)) for (i in c(1,3,5,7)) { plot(survPreds[[i]], estimator = "median", conf.int = TRUE, include.y = TRUE, main = paste("Follow-up time:", round(survPreds[[i]]$last.time, 1)), ylab = "", ylab2 = "") } mtext("log serum bilirubin", side = 2, line = -1, outer = TRUE) mtext("Survival Probability", side = 4, line = -1, outer = TRUE) ## End(Not run)