[R] Warning trying to plot -log(log(survival))
Steven McKinney
smckinney at bccrc.ca
Tue May 12 06:55:55 CEST 2009
Hi John,
Thanks for the data. I put your data in a
plain text file and read it in without any problems.
> jsdf <- read.csv("js.csv")
> jsdf
Time Time30 Died Age Rx
1 3 0.1000000 0 40 2
2 8 0.2666667 1 21 2
3 10 0.3333333 1 18 2
4 12 0.4000000 0 42 2
5 16 0.5333333 1 23 2
...
I can fit the model you specified and obtain the plot without problem.
So this looks like an issue with R 2.8.1 and whatever survival
package version you have. I have R 2.9.0 and survival_2.35-4.
Can you update your versions?
HTH
> fit0 <- coxph(Surv(Time30, Died) ~ strata(Rx) + Age, data = jsdf)
> fit0
Call:
coxph(formula = Surv(Time30, Died) ~ strata(Rx) + Age, data = jsdf)
coef exp(coef) se(coef) z p
Age 0.0611 1.06 0.0257 2.38 0.017
Likelihood ratio test=5.84 on 1 df, p=0.0156 n= 64
> plot(survfit(fit0),fun="cloglog")
### No warning, two curves plotted.
> sessionInfo()
R version 2.9.0 (2009-04-17)
i386-pc-mingw32
locale:
LC_COLLATE=English_United States.1252;LC_CTYPE=English_United
States.1252;LC_MONETARY=English_United
States.1252;LC_NUMERIC=C;LC_TIME=English_United States.1252
attached base packages:
[1] splines stats graphics grDevices utils datasets methods
[8] base
other attached packages:
[1] survival_2.35-4
>
Steven McKinney
> -----Original Message-----
> From: John Sorkin [mailto:jsorkin at grecc.umaryland.edu]
> Sent: Monday, May 11, 2009 8:53 PM
> To: Steven McKinney; r-help at stat.math.ethz.ch
> Subject: RE: [R] Warning trying to plot -log(log(survival))
>
> Time30 is simply Time/30 and reflects months of follow-up.
> A copy of my data follows:
> Time,Time30,Died,Age,Rx
> 3,0.1,0,40,2
> 8,0.266666667,1,21,2
> 10,0.333333333,1,18,2
> 12,0.4,0,42,2
> 16,0.533333333,1,23,2
> 17,0.566666667,1,21,2
> 22,0.733333333,1,13,2
> 64,2.133333333,0,20,2
> 65,2.166666667,0,15,2
> 77,2.566666667,0,34,2
> 82,2.733333333,0,14,2
> 98,3.266666667,0,10,2
> 155,5.166666667,0,27,2
> 189,6.3,0,9,2
> 199,6.633333333,0,19,2
> 247,8.233333333,0,14,2
> 324,10.8,0,23,2
> 356,11.86666667,0,13,2
> 378,12.6,0,34,2
> 408,13.6,0,27,2
> 411,13.7,0,5,2
> 420,14,0,23,2
> 449,14.96666667,0,37,2
> 490,16.33333333,0,37,2
> 528,17.6,0,32,2
> 547,18.23333333,0,32,2
> 691,23.03333333,0,38,2
> 769,25.63333333,0,18,2
> 1111,37.03333333,0,20,2
> 1173,39.1,0,12,2
> 1213,40.43333333,0,12,2
> 1357,45.23333333,0,29,2
> 9,0.3,1,35,1
> 11,0.366666667,1,27,1
> 12,0.4,1,22,1
> 20,0.666666667,1,21,1
> 20,0.666666667,1,30,1
> 22,0.733333333,1,7,1
> 25,0.833333333,1,36,1
> 25,0.833333333,1,38,1
> 25,0.833333333,0,20,1
> 28,0.933333333,1,25,1
> 28,0.933333333,1,28,1
> 31,1.033333333,1,17,1
> 35,1.166666667,1,21,1
> 35,1.166666667,1,25,1
> 46,1.533333333,1,35,1
> 49,1.633333333,1,19,1
> 104,3.466666667,0,27,1
> 106,3.533333333,0,19,1
> 156,5.2,0,15,1
> 218,7.266666667,0,26,1
> 230,7.666666667,0,11,1
> 231,7.7,0,14,1
> 316,10.53333333,0,15,1
> 393,13.1,0,27,1
> 395,13.16666667,0,2,1
> 428,14.26666667,0,3,1
> 469,15.63333333,0,14,1
> 602,20.06666667,0,18,1
> 681,22.7,0,23,1
> 690,23,0,9,1
> 1112,37.06666667,0,11,1
> 1180,39.33333333,0,11,1
>
>
>
>
> John David Sorkin M.D., Ph.D.
> Chief, Biostatistics and Informatics
> University of Maryland School of Medicine Division of Gerontology
> Baltimore VA Medical Center 10 North Greene Street GRECC (BT/18/GR)
> Baltimore, MD 21201-1524
> (Phone) 410-605-7119
> (Fax) 410-605-7913 (Please call phone number above prior to faxing)
>
> >>> Steven McKinney <smckinney at bccrc.ca> 5/11/2009 10:04 PM >>>
> Hi John,
>
> I can't reproduce your case with built-in data sets.
>
> Your data set is small. Can you show the data for the variables
> involved in your example?
> (Time30, Died, Rx, Age)
>
>
>
>
>
> Steven McKinney, Ph.D.
>
> Statistician
> Molecular Oncology and Breast Cancer Program British Columbia Cancer
> Research Centre
>
> email: smckinney at bccrc.ca
> tel: 604-675-8000 x7561
>
> BCCRC
> Molecular Oncology
> 675 West 10th Ave, Floor 4
> Vancouver B.C.
> V5Z 1L3
>
> Canada
>
>
>
>
>
>
>
> > -----Original Message-----
> > From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-
> > project.org] On Behalf Of John Sorkin
> > Sent: Monday, May 11, 2009 12:19 PM
> > To: r-help at stat.math.ethz.ch
> > Subject: [R] Warning trying to plot -log(log(survival))
> >
> > windows xp
> > R 2.8.1
> >
> > I am trying to plot the -log(log(survival)) to visually test the
> > proportional hazards assumption of a Cox regression. The plot, which
> > should give two lines (one for each treatment) gives only one line
> and
> > a warning message. I would appreciate help getting two lines, and an
> > explanation of the warning message. My problem may the that I have
> very
> > few events in one of my strata, but I don't know.
> > Thanks,
> > John
> >
> > fit0<-coxph(Surv(Time30,Died)~strata(Rx)+Age,data=GVHDdata)
> > plot(survfit(fit0),fun="cloglog")
> >
> >
> > WARNING: Warning in xy.coords (x, y, xlabel, ylabel, log) :
> > 2 x values <=0 omitted from logarithmic plot
> >
> >
> >
> > > print(survfit(fit0),fun="cloglog")
> > Call: survfit.coxph(object = fit0)
> >
> > n events median 0.95LCL 0.95UCL
> > Rx=0 31 5 Inf Inf Inf
> > Rx=1 32 15 Inf 1.17 Inf
> >
> > John David Sorkin M.D., Ph.D.
> > Chief, Biostatistics and Informatics
> > University of Maryland School of Medicine Division of Gerontology
> > Baltimore VA Medical Center 10 North Greene Street GRECC (BT/18/GR)
> > Baltimore, MD 21201-1524
> > (Phone) 410-605-7119
> > (Fax) 410-605-7913 (Please call phone number above prior to faxing)
> >
> > Confidentiality Statement:
> > This email message, including any attachments, is for
> > th...{{dropped:6}}
> >
> > ______________________________________________
> > R-help at 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.
>
> Confidentiality Statement:
> This email message, including any attachments, is for ...{{dropped:6}}
More information about the R-help
mailing list