FR: More documentation on aareg

[iago-pssjd]

Dear Terry,

I was trying to understand aareg output/inputs through survival documentation beyond ?aareg, coming to

The timereg package is a much more comprehensive implementation of the Aalen model, so this document will say little about aareg

in survival vignette. Unfortunately, the output of timereg::aalen does not seem to have much to do with that of aareg and after a quite deep search on the web for comparisons between the 2 functions I couldn't find anything. For example,

• where I can found the outputs of both functions in the other ones?
• how dfbeta relates to the input (as Chisq and other values change if dfbeta = FALSE) and is this option available somehow in timereg::aalen?
• does Chisq correspond to some value of timereg::aalen? Which test is it evaluating? (I mean, in the Aalen's book Survival and Event History Analysis I found many test statistics "chi-squared distributed", with which of them would it correspond?)
• should I look maybe to the test.statistic value of aareg?

library(survival)
afit <- aareg(
  Surv(time, status) ~ age + sex + ph.ecog,
  data = lung,
  dfbeta = TRUE
)

afit

Call:
aareg(formula = Surv(time, status) ~ age + sex + ph.ecog, data = lung, 
    dfbeta = TRUE)

  n=227 (1 observation deleted due to missingness)
    136 out of 138 unique event times used

              slope      coef se(coef) robust se     z        p
Intercept  5.05e-03  5.87e-03 4.74e-03   0.00477  1.23 0.219000
age        4.01e-05  7.15e-05 7.23e-05   0.00007  1.02 0.307000
sex       -3.16e-03 -4.03e-03 1.22e-03   0.00123 -3.28 0.001030
ph.ecog    3.01e-03  3.67e-03 1.02e-03   0.00102  3.62 0.000299

Chisq=22.84 on 3 df, p=4.36e-05; test weights=aalen


summary(afit)

$table
                  slope          coef     se(coef)    robust se         z            p
Intercept  5.048983e-03  5.868616e-03 4.739162e-03 4.771021e-03  1.230055 0.2186766165
age        4.005089e-05  7.149015e-05 7.228889e-05 6.996847e-05  1.021748 0.3069001400
sex       -3.164485e-03 -4.030555e-03 1.217949e-03 1.227954e-03 -3.282333 0.0010295209
ph.ecog    3.009913e-03  3.673470e-03 1.016785e-03 1.015845e-03  3.616171 0.0002989931

$test
[1] "aalen"

$test.statistic
 Intercept        age        sex    ph.ecog 
  1.901744 108.155068 -19.531696  33.158152 

$test.var
            b0                                  
b0    2.358499  -151.80330 -3.715147    2.157013
   -151.803299 11960.36511 16.697824 -277.142084
     -3.715147    16.69782 34.834385   -5.183630
      2.157013  -277.14208 -5.183630   84.233750

$test.var2
            [,1]        [,2]       [,3]        [,4]
[1,]    2.390315  -149.43604  -4.237668    1.406427
[2,] -149.436042 11204.85060  69.078254 -169.293859
[3,]   -4.237668    69.07825  35.409086  -12.726907
[4,]    1.406427  -169.29386 -12.726907   84.078078

$chisq
         [,1]
[1,] 22.84047

$n
[1] 227 136 138

attr(,"class")
[1] "summary.aareg"


out <- aalen(
    Surv(time, status) ~ age + sex + ph.ecog,
    data = lung
)
summary(out)

Additive Aalen Model 

Test for nonparametric terms 

Test for non-significant effects 
            Supremum-test of significance p-value H_0: B(t)=0
(Intercept)                           Inf                   0
age                                   Inf                   0
sex                                   Inf                   0
ph.ecog                               Inf                   0

Test for time invariant effects 
                  Kolmogorov-Smirnov test p-value H_0:constant effect
(Intercept)                        2.5200                       0.338
age                                0.0468                       0.205
sex                                0.6830                       0.358
ph.ecog                            0.4970                       0.401
                    Cramer von Mises test p-value H_0:constant effect
(Intercept)                       436.000                       0.536
age                                 0.212                       0.278
sex                                84.600                       0.221
ph.ecog                            40.100                       0.337

   
   
  Call: 
aalen(formula = Surv(time, status) ~ age + sex + ph.ecog, data = lung)

Many thanks!

[iago-pssjd]

By the way, as you can see in the summary(afit) output, it seems that S3method(print, summary.aareg) should be included in the NAMESPACE file.

Thanks!