[Rd] Adding a survival object to a data frame (PR#10510)

Prof Brian Ripley ripley at stats.ox.ac.uk
Thu Dec 13 12:09:46 CET 2007


Apparently this was Surv from package Design.
So the bug is in contributed package Design, and nothing to do with 
R-bugs.

On Wed, 12 Dec 2007, Peter Dalgaard wrote:

> ripley at stats.ox.ac.uk wrote:
>> Your example is not reproducible without 'library(survival)'.
>> When I include that, I get
>>
>>
>>> head(D,20)
>>>
>>     stime status  surv
>> 1    176   TRUE  176
>> ...
>>
>> Objects of class "Surv" are from the contributed package survival, and you
>> need that attached to deal with them properly.
>>
>>
>>
> ... but even if you detach it, you do not get the symptoms shown:
>
>> head(D,5)
>   stime status surv.time surv.status
> 1    176   TRUE       176           1
> 2     67   TRUE        67           1
> 3    432   TRUE       432           1
> 4     77   TRUE        77           1
> 5    275   TRUE       275           1
>
>
>> On Wed, 12 Dec 2007, edward.m at psu.ac.th wrote:
>>
>>
>>> Full_Name: Edward McNeil
>>> Version: 2.6.1
>>> OS: Windows
>>> Submission from: (NULL) (203.170.234.5)
>>>
>>>
>>> I want to show students how the survival object looks like in R.
>>> Reproducible example:
>>>
>>> library(MASS)
>>> data(Aids2)
>>> attach(Aids2)
>>> status <- status=="D"
>>> stime <- death-diag
>>> surv <- Surv(stime, status)
>>> D <- data.frame(stime, status, surv)
>>> head(D,20)
>>>   stime status x..i..
>>> 1    176   TRUE   176
>>> 2     67   TRUE    67
>>> 3    432   TRUE   432
>>> 4     77   TRUE    77
>>> 5    275   TRUE   275
>>> 6    373   TRUE   373
>>> 7    389   TRUE   389
>>> 8   1027   TRUE  1027
>>> 9    492   TRUE   492
>>> 10   434   TRUE   434
>>> 11    16   TRUE    16
>>> 12   308   TRUE   308
>>> 13    92   TRUE    92
>>> 14   265   TRUE   265
>>> 15  1052  FALSE  1052+
>>> 16   132   TRUE   132
>>> 17   527   TRUE   527
>>> 18   581  FALSE   581+
>>> 19   511  FALSE   511+
>>> 20   151   TRUE   151
>>>
>>> detach(Aids2)
>>>
>>> The 'surv' column is strangely labelled 'x..i..'.


-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595



More information about the R-devel mailing list