\(\hat q_j = d_j/(n_j - w_j/2)\) and
\(\hat S_j = \prod_{k \le j}(1 - \hat q_k)\). The \(w_j/2\) is
the method's only real content: withdrawals are assumed uniform
within the interval, so each contributes half an interval of
exposure on average. Kaplan-Meier is the limit as intervals
shrink and should be preferred whenever exact times exist.
Mirrors morie.fn.smplts.
Usage
morie_actuarial_lifetable(intervals, entered, died, withdrawn = NULL)
Arguments
- intervals
increasing interval boundaries, length J+1.
- entered
number entering each interval.
- died
deaths in each interval.
- withdrawn
withdrawals in each interval.
Value
list: intervals, q, survival, effective_n,
actuarial_correction, assumes, prefer, J, method.
Examples
morie_actuarial_lifetable(c(0, 1, 2), c(100, 70), c(20, 20),
c(10, 10))$q
#> [1] 0.2105263 0.3076923
