plot.lwy.test {surv2sample} | R Documentation |
This function plots an observed test process along with simulated paths. Such a plot may be used to visually assess the hypothesis.
## S3 method for class 'lwy.test': plot(x, lwds = c(3, 1), cols = c("black", "grey"), ltys = c(1, 1), ...)
x |
an object of class "lwy.test" . |
lwds |
a vector of length 2, giving the line width for the observed process and for simulated processes. |
cols |
a vector of length 2, giving the line colour for the observed process and for simulated processes. |
ltys |
a vector of length 2, giving the line type for the observed process and for simulated processes. |
... |
other standard parameters for plotting. |
In survival analysis, p-values of Kolmogorov–Smirnov and other tests based on a test process may be approximated by the martingale-based simulation technique originally proposed by Lin, Wei and Ying (1993). By plotting the observed test process along with a suitable number of simulated paths, one may visually assess the validity of the hypothesis.
This function is a general plot method for such tests. Results
of such tests are objects with class attribute "lwy.test"
,
at least containing sorted times in x$time
, the observed process
in x$test.process
and a number (x$nsim.plot
) of simulated
processes in x$test.process.sim
. Results of functions listed in
‘See Also’ below are handled by this plot method.
David Kraus (http://www.davidkraus.net/)
Lin, D. Y., Wei, L. J. and Ying, Z. (1993) Checking the Cox model with cumulative sums of martingale-based residuals. Biometrika 80, 557–572.
surv2.ks
, cif2.ks
, proprate2.ks
## gastric cancer data data(gastric) ## Kolmogorov--Smirnov test of equal survival distributions ## test process = difference of Nelson--Aalen estimates ## plot the observed test process and simulated paths plot(surv2.ks(Surv(gastric$time, gastric$event), gastric$treatment, approx="mart"))