Hi All,
I want to get the OS and EFS curve to compare two groups by log-rank
test. I can draw the two curves by "survfit" but I couldn't find
"log-rank" test in the choices of argument. I also looked at the
"survdiff" which does have "log-rank" test, but it seems like no plots
can be drawn from it.
Could you please help me with it?
The attached are the data and code.
Thank you!
UPN SURVEFP STATEFP SURVP STATP STATNRMP STATREP DXCODE TYPE
318 41 1 41 1 1 0 2 1
492 2078 1 2078 1 1 0 2 2
625 128 1 128 1 1 0 1 2
631 656 1 1548 1 0 1 2 2
637 16 1 16 1 1 0 2 2
652 17 1 17 1 1 0 1 2
759 66 1 156 1 0 1 1 2
826 52 1 52 1 1 0 1 2
918 121 1 289 1 0 1 2 1
985 152 1 166 1 0 1 1 2
1033 85 1 103 1 0 1 1 2
1222 799 1 967 1 0 1 1 2
1225 182 1 639 1 0 1 1 2
1228 2575 0 2575 0 0 0 2 1
1319 106 1 106 1 1 0 1 2
1321 962 1 2622 0 0 1 2 1
1370 61 1 68 1 0 1 2 1
1468 2212 0 2212 0 0 0 2 2
1528 245 1 456 1 0 1 1 2
1634 1324 1 2074 1 0 1 1 2
1687 769 1 1461 1 0 1 1 2
1810 104 1 173 1 0 1 1 2
1824 92 1 133 1 0 1 1 2
1867 285 1 285 1 1 0 2 2
1889 1525 0 1525 0 0 0 1 2
1893 1529 0 1529 0 0 0 2 2
1895 1009 0 1009 0 0 0 1 2
1917 334 1 334 1 1 0 1 2
1958 38 1 38 1 1 0 2 2
1972 197 1 197 1 1 0 1 2
1979 1275 0 1275 0 0 0 1 2
2009 82 1 223 1 0 1 2 1
2039 1 1 124 1 0 1 1 2
2061 91 1 91 1 1 0 1 2
2114 754 0 754 0 0 0 1 2
2115 332 0 332 0 0 0 2 2
2117 85 1 98 1 0 1 2 2
2119 41 1 41 1 1 0 2 2
2135 174 1 512 1 0 1 2 1
2141 679 1 679 1 1 0 1 2
2159 722 0 722 0 0 0 1 1
2167 733 0 733 0 0 0 2 2
2173 643 0 643 0 0 0 2 2
2179 699 0 699 0 0 0 2 2
2236 563 0 563 0 0 0 1 2
2250 386 0 386 0 0 0 1 1
2271 25 1 25 1 1 0 1 2
2295 396 0 396 0 0 0 1 1
2304 295 1 295 1 1 0 2 2
2307 348 0 348 0 0 0 2 1
2311 77 1 107 1 0 1 2 2
2340 174 1 317 0 0 1 1 1
2360 37 0 37 0 0 0 1 1
2361 263 0 263 0 0 0 2 1
dat<-read.table("E://my file/BC project/Project II/KM
estimates/data.txt",header=TRUE,sep="\t")
attach(dat)
## OS for whole group
# NO CI
OS_whole<-survfit(Surv(SURVP/365.25,STATP)~1,type="kaplan-meier")
plot(OS_whole,main="OS for whole
group",conf.int=F,xlab="years",ylab="probability")
# 95% cI
OS_whole<-survfit(Surv(SURVP/365.25,STATP)~1,type="kaplan-meier",conf.type="log-log")
plot(OS_whole,main="OS for whole
group",conf.int=T,xlab="years",ylab="probability")
mtext("with 95% confidence interval",side=3)
## EFS for whole group
# NO CI
EFS_whole<-survfit(Surv(SURVEFP/365.25,STATEFP)~1,type="kaplan-meier")
plot(EFS_whole,main="EFS for whole
group",conf.int=F,xlab="years",ylab="probability")
# 95% CI
EFS_whole<-survfit(Surv(SURVEFP/365.25,STATEFP)~1,type="kaplan-meier",conf.type="log-log")
plot(EFS_whole,main="EFS for whole
group",conf.int=T,xlab="years",ylab="probability")
mtext("with 95% confidence interval",side=3)
## OS for group split into allo/auto with log-rank comparison
OS_aa<-survdiff(Surv(SURVP,STATP)~TYPE,rho=0)______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
