newllk {PHInfiniteEstimates} | R Documentation |
Proportional hazards partial likelihood, using Breslow method for ties, excluding some observations.
Description
This function implements the approximate conditional inferential approach of Kolassa and Zhang (2019) to proportional hazards regression.
Usage
newllk(
beta,
fit,
exclude = NULL,
minus = FALSE,
keeponly = NULL,
justd0 = FALSE,
cc1 = 0
)
Arguments
beta |
parameter vector. |
fit |
Output from a Cox PH regression, with x=TRUE and y=TRUE |
exclude |
data set with stratum and patient number to exclude. |
minus |
logical flag to change sign of partial likelyhood |
keeponly |
variables to retain. Keep all if this is null or NA. |
justd0 |
logical variable, indicating whether to calculate only the function value and skip derivatives. |
cc1 |
Continuity correction for first component of the score. |
Value
a list with components
d0 partial likelihood
d1 first derivative vector
d2 second derivative matrix
References
Kolassa JE, Zhang J (2019). https://higherlogicdownload.s3.amazonaws.com/AMSTAT/fa4dd52c-8429-41d0-abdf-0011047bfa19/UploadedImages/NCB_Conference/Presentations/2019/kolassa_toxslides.pdf. Accessed: 2019-07-14.