summary.crr {cmprsk} | R Documentation |
Generate and print summaries of crr output
## S3 method for class 'crr'
summary(object, conf.int = 0.95, digits =
max(options()$digits - 5, 2), ...)
## S3 method for class 'summary.crr'
print(x, digits = max(options()$digits - 4, 3), ...)
object |
An object of class crr (output from the crr function) |
conf.int |
the level for a two-sided confidence interval on the coeficients. Default is 0.95. |
digits |
In |
... |
Included for compatibility with the generic functions. Not currently used. |
x |
An object of class summary.crr (output from the summary method for crr) |
The summary method calculates the standard errors, subdistribution hazard ratios z-scores, p-values, and confidence intervals on the hazard ratios. The print method prints a fairly standard format tabular summary of the results.
The pseudo likelihood ratio test in the printed output is based on the difference in the objective function at the global null and at the final estimates. Since this objective function is not a true likelihood, this test statistic is not asymptotically chi-square.
summary.crr
returns a list of class summary.crr, which
contains components
call |
The call to crr |
converged |
TRUE if the iterative algorithm converged |
n |
The number of observations used in fitting the model |
n.missing |
The number of observations removed by |
loglik |
The value of the negative of the objective function (the pseudo log likelihood at convergence |
coef |
A matrix giving the estimated coefficients, hazard ratios, standard errors, z-scores, and p-values |
conf.int |
A matrix giving the estimated hazard ratios, inverse hazard ratios and lower and upper confidence limits on the hazard ratios |
logtest |
Twice the difference in log pseudo likelihood values |
The summary and print.summary methods were provided by Luca Scrucca
## see examples in the crr help file