[R] coxph() - unexpected result using Crawley's seedlings data (The R Book)

Jacob Brogren jacob at brogren.nu
Tue Jun 28 17:04:20 CEST 2011


Hi,

sorry about that; here is the full output - data set, structure, model and result.

Cheers

Jacob

> seedlings
      cohort death gapsize status
1  September     7  0.5889      1
2  September     3  0.6869      1
3  September    12  0.1397      1
4  September     1  0.1921      1
5  September     4  0.2798      1
6  September     2  0.2607      1
7  September     6  0.9467      1
8  September     6  0.6375      1
9  September     8  0.1527      1
10 September     3  0.8237      1
11 September     1  0.5979      1
12 September     1  0.2914      1
13 September     3  0.5053      1
14 September     5  0.4714      1
15 September     2  0.6041      1
16 September     8  0.8812      1
17 September     4  0.8416      1
18 September     1  0.0577      1
19 September     2  0.9034      1
20 September     2  0.4348      1
21 September     7  0.9878      1
22 September    11  0.1486      1
23 September    14  0.5003      1
24 September     1  0.8507      1
25 September    10  0.8187      1
26 September    14  0.0291      1
27 September     1  0.3785      1
28 September     4  0.8384      1
29 September     2  0.8351      1
30 September     2  0.9674      1
31   October     1  0.6943      1
32   October     1  0.2591      1
33   October     2  0.7397      1
34   October     2  0.4663      1
35   October    14  0.9115      1
36   October     5  0.1750      1
37   October     1  0.5628      1
38   October     8  0.2681      1
39   October     5  0.6967      1
40   October     2  0.7020      1
41   October     4  0.7971      1
42   October     3  0.4047      1
43   October     5  0.0498      1
44   October    10  0.0364      1
45   October     9  0.4080      1
46   October     1  0.6226      1
47   October    11  0.3002      1
48   October     3  0.8111      1
49   October    21  0.4894      1
50   October     1  0.0375      1
51   October     4  0.2560      1
52   October     9  0.2168      1
53   October     8  0.7437      1
54   October     1  0.9082      1
55   October     3  0.9496      1
56   October     9  0.1040      1
57   October     9  0.8691      1
58   October    16  0.9502      1
59   October     6  0.0790      1
60   October     1  0.5658      1
> str(seedlings)
'data.frame':	60 obs. of  4 variables:
 $ cohort : Factor w/ 2 levels "October","September": 2 2 2 2 2 2 2 2 2 2 ...
 $ death  : int  7 3 12 1 4 2 6 6 8 3 ...
 $ gapsize: num  0.589 0.687 0.14 0.192 0.28 ...
 $ status : num  1 1 1 1 1 1 1 1 1 1 ...
> model1 <- coxph(Surv(death,status)~strata(cohort)*gapsize,data=seedlings)
> summary(model1)
Call:
coxph(formula = Surv(death, status) ~ strata(cohort) * gapsize, 
    data = seedlings)

  n= 60, number of events= 60 

                                            coef exp(coef)  se(coef)      z Pr(>|z|)
gapsize                                -0.001893  0.998109  0.593372 -0.003    0.997
gapsize:strata(cohort)cohort=September  0.717407  2.049112  0.860807  0.833    0.405

                                       exp(coef) exp(-coef) lower .95 upper .95
gapsize                                   0.9981      1.002    0.3120     3.193
gapsize:strata(cohort)cohort=September    2.0491      0.488    0.3792    11.074

Rsquare= 0.022   (max possible= 0.993 )
Likelihood ratio test= 1.35  on 2 df,   p=0.5097
Wald test            = 1.32  on 2 df,   p=0.5178
Score (logrank) test = 1.33  on 2 df,   p=0.514

28 jun 2011 kl. 15.48 skrev Robert A LaBudde:

> Did you create the 'status' variable the way indicated on p. 797?
> 
> Frequently with Surv() it pays to use syntax such as Surv(death, status==1) to make a clear logical statement of what is an event (status==1) vs. censored.
> 
> PS. Next time include head(seedlings) and str(seedlings) to make clear what you are using as data.
> 
> 
> At 06:51 AM 6/28/2011, Jacob Brogren wrote:
>> Hi,
>> 
>> I ran the example on pp. 799-800 from Machael Crawley's "The R Book" using package survival v. 2.36-5, R 2.13.0 and RStudio 0.94.83. The model is a Cox's Proportional Hazards model. The result was quite different compared to the R Book. I have compared my code to the code in the book but can not find any differences in the function call. My results are attached as well as a link to the results presented in the book (link to Google Books).
>> 
>> When running the examples on pp. 797-799 I can't detect any differences in results so I don't think there are errors in the data set or in the creation of the status variable.
>> 
>> ---------------------------------
>> Original from the R Book:
>> http://books.google.com/books?id=8D4HVx0apZQC&lpg=PA799&ots=rQgd_8ofeS&dq=r%20coxph%20crawley&pg=PA799#v=onepage&q&f=false
>> 
>> ---------------------------------
>> My result:
>> > summary(model1)
>> Call:
>> coxph(formula = Surv(death, status) ~ strata(cohort) * gapsize,
>>    data = seedlings)
>> 
>>  n= 60, number of events= 60
>> 
>>                                            coef exp(coef)  se(coef)      z Pr(>|z|)
>> gapsize                                -0.001893  0.998109  0.593372 -0.003    0.997
>> gapsize:strata(cohort)cohort=September  0.717407  2.049112  0.860807  0.833    0.405
>> 
>>                                       exp(coef) exp(-coef) lower .95 upper .95
>> gapsize                                   0.9981      1.002 0.3120     3.193
>> gapsize:strata(cohort)cohort=September    2.0491      0.488 0.3792    11.074
>> 
>> Rsquare= 0.022   (max possible= 0.993 )
>> Likelihood ratio test= 1.35  on 2 df,   p=0.5097
>> Wald test            = 1.32  on 2 df,   p=0.5178
>> Score (logrank) test = 1.33  on 2 df,   p=0.514
>> 
>> Anyone have an idea why this is occurring?
>> 
>> Kind Regards
>> 
>> Jacob
>> ______________________________________________
>> 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.
> 
> ================================================================
> Robert A. LaBudde, PhD, PAS, Dpl. ACAFS  e-mail: ral at lcfltd.com
> Least Cost Formulations, Ltd.            URL: http://lcfltd.com/
> 824 Timberlake Drive                     Tel: 757-467-0954
> Virginia Beach, VA 23464-3239            Fax: 757-467-2947
> 
> "Vere scire est per causas scire"
> ================================================================
> 



More information about the R-help mailing list