On May 27, 2010, at 12:42 AM, Wu Gong wrote:
I don't know if my understanding of P is right.
P ?= (the number of lives at the end of the interval)/(the number of
lives
at the beginning of the interval)
In standard actuarial parlance one would use lower case p_x for that
quantity and would reserve upper case P_x for the cumulative survival
fraction. The cumulative product of the p_x', P_x, with appropriate
corrections for censoring, was rebranded in biostatistical circles as
the Kaplan-Meier estimator.
### Compute proportion of a cohort that survives through the interval
### The formula is P0=L1/LO
## Original data is a vector of death days
example=as.vector(c(8,2,1,5,6,7,7,0,8,10,13,8,11,11,11,2,7,1,5,6,8,6))
## Count the frenquency of each death day
## Add more levels to fulfill days range
tb <- table(factor(example,levels=seq(0, max(example))))
## Output
iC <- data.frame("x"=names(tb),
"Kx"=sum(tb)-cumsum(tb),
"Dx"=tb[],
"Lx"=iC$Kx/sum(tb),
"Px"=iC$Kx/(iC$Kx+tb[]))
In a fresh session that R code does not succeed, since the data.frame
function does not handle recursive references.
-----
A R learner.
David Winsemius, MD
West Hartford, CT
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
