Consult a local statistician. You need to understand the concept of
contrasts for categorical variables, which is a basic statistical
issue, not an R issue. How R handles contrasts is explained (tersely)
in ?contrast.
Cheers,
Bert
Bert Gunter
Genentech Nonclinical Biostatistics
(650) 467-7374
"Data is not information. Information is not knowledge. And knowledge
is certainly not wisdom."
Clifford Stoll
On Fri, Apr 10, 2015 at 10:23 AM, CHIRIBOGA Xavier
<xavier.chiriboga at unine.ch> wrote:> Dear members,
>
>
>
> I did a survival analysis, Weibul fits better for my data: this is the
"head" of my data
>
>
>
> start stop state soil volatile hours replicate
> 1 0 48 1 ca nc 48 3
> 2 0 48 1 ca nc 48 4
> 3 0 48 1 ca nc 48 5
> 4 0 48 1 ca nc 48 8
> 5 0 60 1 ca c 60 1
> 6 0 60 1 ca c 60 1
>
>
>
>
>
> I have 4 soils: ca, cs, cl and sand
>
>
>
> and I have 2 volatiles: nc and c
>
>
>
> However, when I get the stimates my soil "ca" does not appear and
the volatile "c" neither
>
>
>
> Call:
> survreg(formula = Surv(hours, state) ~ soil * volatile, data = surgal,
> dist = "weibull")
> Value Std. Error z p
> (Intercept) 4.3605 0.0417 104.51 0.00e+00
> soilcl -0.0784 0.0590 -1.33 1.84e-01
> soilcs -0.3865 0.0567 -6.81 9.59e-12
> soilsand -0.1913 0.0579 -3.30 9.61e-04
> volatilenc -0.1600 0.0579 -2.76 5.74e-03
> soilcl:volatilenc 0.1320 0.0823 1.60 1.09e-01
> soilcs:volatilenc 0.2679 0.0812 3.30 9.64e-04
> soilsand:volatilenc 0.0821 0.0809 1.01 3.10e-01
> Log(scale) -1.4788 0.0442 -33.46 1.73e-245
>
> Scale= 0.228
>
> Weibull distribution
> Loglik(model)= -1036.3 Loglik(intercept only)= -1065
> Chisq= 57.41 on 7 degrees of freedom, p= 5e-10
> Number of Newton-Raphson Iterations: 5
> n= 257
>
>
>
> Anyone knows why? or what does it mean? how should I interpret the values
in the table?
>
>
>
> Thanks for your help,
>
>
>
> Xavier
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> 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.