[R] Survival Plots by Strata

[Bret Collier]

Thomas,

Thank you for the response.  I will post what I come up with.

Bret

>>> Thomas Lumley <tlumley at u.washington.edu> 03/08/06 9:16 AM >>>

I don't see any easy way to do this.  I think you may have to do the 
plotting yourself, based on the code in plot.survfit.

        -thomas


On Wed, 8 Mar 2006, Bret Collier wrote:

> All,
> I am struggling to create a survival plot using LTRC data for each
year
> of a 10 year period.
>
> I have a set of individuals (birds) where 'entry' is the day of the
> year (1-365) they are released (let out of pens) into the wild (2
year
> data snip below).  'Entry'  (e.g., day of year the first bird is
> released for each year) is highly variable, ranging from 48 to >250.
> When I create a plot using the below code I would like to remove the
> 'solid' line which represent S_hat=1 out to the LC point for each
year
> and instead have the survival curves formatted in more of a
'hanging'
> style, where the LC day (e.g., min(Entry for each year)) is where
each
> curve begins at S_hat=1 for each year (and does not extend back to
the
> y-axis)?  I could not find anything on this in the archives or MASS
or
> Survival Analysis using S?  Anyone have a suggestion on where to
look?
>
> TIA, Bret
>
>
> #Code snip for R email.
> apc.coxfit1<-coxph(Surv(Entry, Exit, Fate)~Sex + Agerelease +
> Dayrelease + strata(Year), data=mydat)
> coxfit.apc<-survfit(apc.coxfit1)
> coxfit.apc
> plot(survfit(apc.coxfit1), conf.int=F, log=T, lty=c(1:2),
col=c(1:2),
> xlim=c(205, 800)) #not run--first entry for this example is day 205
for
> 1996, 259 for 1997
>
>> mydat
> ID	Year	Dayrelease
> Agerelease	Survivorship	Entry	Exit	Fate	Sex
>
16240	1996	205	95	164	205	369	1	0
>
16319	1996	205	88	140	205	345	1	0
>
16378	1996	248	108	100	248	348	1	0
>
20383	1996	241	98	204	241	445	1	0
>
16324	1996	219	90	227	219	446	1	0
>
16327	1996	219	90	497	219	716	1	0
>
20373	1996	241	114	413	241	654	1	0
>
20374	1996	241	111	211	241	452	1	0
>
16241	1996	205	95	234	205	439	1	1
>
16321	1996	219	90	118	219	337	1	1
>
16323	1996	219	90	180	219	399	1	1
>
20375	1996	241	103	268	241	509	1	1
>
20384	1996	241	98	299	241	540	1	1
>
20390	1996	241	93	204	241	445	1	1
>
20393	1996	241	88	208	241	449	1	1
>
16313	1996	248	122	512	248	760	0	1
>
20378	1996	241	103	236	241	477	0	1
>
20381	1996	241	101	329	241	570	0	1
>
16328	1996	219	90	224	219	443	0	1
>
16827	1997	259	127	52	259	311	1	0
>
16828	1997	259	127	216	259	475	1	0
>
16831	1997	303	171	19	303	322	1	0
>
20466	1997	289	149	31	289	320	1	0
>
20469	1997	289	149	199	289	488	1	0
>
20483	1997	289	134	18	289	307	1	0
>
16807	1997	259	137	223	259	482	1	0
>
16809	1997	259	137	1	259	260	1	0
>
16819	1997	259	131	237	259	496	1	0
>
16829	1997	303	171	7	303	310	1	0
>
20440	1997	289	161	7	289	296	1	0
>
20470	1997	289	148	257	289	546	1	0
>
20478	1997	289	143	12	289	301	1	0
>
16817	1997	259	130	85	259	344	1	0
>
20454	1997	289	154	4	289	293	1	1
>
20459	1997	289	153	335	289	624	1	1
>
20460	1997	289	153	118	289	407	1	1
>
20465	1997	289	150	31	289	320	1	1
>
20473	1997	289	147	65	289	354	1	1
>
20484	1997	289	133	58	289	347	1	1
>
20489	1997	289	130	56	289	345	1	1
>
16808	1997	259	137	137	259	396	0	1
>
16810	1997	303	181	64	303	367	0	1
>
16816	1997	259	130	1	259	260	0	1
>
20471	1997	289	147	334	289	623	0	1
>
16826	1997	259	127	20	259	279	0	1
>
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help 
> PLEASE do read the posting guide!
http://www.R-project.org/posting-guide.html 
>

Thomas Lumley			Assoc. Professor, Biostatistics
tlumley at u.washington.edu	University of Washington, Seattle