Description Usage Arguments Details Value
Computes nonparametric estimates of the cumulative distribution function using maximum likelihood. If possible, estimation this is done with a generalized version of the Kaplan-Meier estimate, and otherwise, by using Turnbull's e-m algorithm
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 |
y |
Numeric matrix |
ny |
Number of columns in |
codes |
Integer vector of censor codes (see Details) |
weight |
Vector of observation weights or multiplicities |
ty |
Numeric matrix containing lower and upper truncation limits |
nty |
Number of columns in ty |
tcodes |
Integer vector of truncation codes (see details) |
n |
Number of rows in y |
nstart |
If |
dscrat |
Numeric scratch vector (length = 3 * n + 2) |
scrat |
Numeric scratch array (length = max((7 * n), maxmsd * (maxmsd - 1) / 2)) |
iscrat |
Integer scratch array (length = 6 * n + 4) |
maxit |
Maximum number of iterations for s-c algorithm |
tol |
Desired estimation accuracy (0 < |
maxmsd |
Maximum |
iprint |
Print level for debug dump |
0dummy observation
1exact failure time
2right censored observation
3left censored observation
4interval or group censored observation
5exact failure time recoded as a small interval
1no truncation
2right truncated observation
3left truncated observation
4interval truncated observation
iprint = 0
for no debug output, if > 0 dump setup and every iprint iteration,
iprint = 1
gives the maximum amount of output, iprint = 10
is a good choice to debug
A list
of length 6
pNumeric vector of lower limits of intervals for the cdf estimate (length = n + 1)
qNumeric vector of upper limits of intervals for the cdf estimate (length = n + 1)
probNumeric vector of cdf estimates corresponding to intervals p
and q
mActual length of p
, q
, and prob
(depends on data)
pchmaxMax change in last iteration (if greater than tol, 0.0 if Kaplan-Meier was used)
ierError code (see details)
0 no error 1 n <= 0 on input 2 ny not equal to 1 or 2 3 nty not between 0 and 2 4 tol outside range (0, 0.1) 6 censor code out of range 0 to 4 7 y(i,1) != y(i,2) in a type 1, 2, or 3 observation 8 y(i,1) < y(i,2) in type 4 obs 9 ny=1 but type 4 obs found 10 tcode outside range 1 to 4 11 tcode = 1, 2, or 3 but tyl != tyu 12 backwards truncation interval 12 ty(i,1) < ty(i,2) in type 4 obs 13 only 1 col of truncation values, code 4 found 14 not enough data to estimate distribution (e.g., all right censored observations) 15 nstart>0 does not agree with computed m 16 observation not within the truncation interval *21 product-limit estimate could not be computed directly and maxmsd was too small to allow the full information matrix to be computed an approximation was computed under the assumption that the individual hazard estimates are uncorrelated *22 information matrix not positive definite *23 only one non-zero s probability
* signifies warning message only
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.