r - 为什么 R 不 relevel？

Question

请在下面找到我的数据“w”。

我有协变量 w$WHO，它具有三个级别:w$WHO==1、w$WHO==2 和 w$WHO==3

我想重新升级，所以 w$WHO==1 被设置为引用。

我试过了

w$WHO <- factor(w$WHO)
w$WHO <- relevel(w$WHO, ref=1)

和

w$WHO <- relevel(w$WHO, ref="1")

我的脚本是

library(rms)
d <- datadist(w)
options(datadist="d")

model <- cph(Surv(rfs,recurrence)~age + WHO,data=w)
summary(model)

如您所见，调整后的模型选择w$WHO==2作为引用。

            Effects              Response : Surv(rfs, recurrence) 

 Factor        Low    High   Diff.  Effect   S.E.     Lower 0.95 Upper 0.95
 age           48.545 68.907 20.362  0.28228 0.093283  0.099454   0.46512  
  Hazard Ratio 48.545 68.907 20.362  1.32620       NA  1.104600   1.59220  
 WHO - 1:2      2.000  1.000     NA -0.56706 0.156850 -0.874490  -0.25963  
  Hazard Ratio  2.000  1.000     NA  0.56719       NA  0.417080   0.77134  
 WHO - 3:2      2.000  3.000     NA  0.69360 0.152910  0.393910   0.99330  
  Hazard Ratio  2.000  3.000     NA  2.00090       NA  1.482800   2.70010  

这是我的数据

My data

w <- structure(list(age = c(54.36164384, 74.91232877, 64.98356164, 
60.56712329, 57.61369863, 45.85205479, 78.47123288, 59.95616438, 
57.4739726, 25.12876712, 56.61917808, 61.10136986, 58.74246575, 
62.56438356, 55.81917808, 30.83013699, 63.11232877, 56.29863014, 
47.96986301, 40.53424658, 49.9890411, 47.75616438, 40.83835616, 
42.02191781, 49.85205479, 55.05479452, 59.33424658, 71.89589041, 
60.30410959, 50.24383562, 41.3260274, 33.4, 73.27945205, 67.45753425
), WHO = c(3L, 1L, 2L, 2L, 1L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 3L, 
2L, 3L, 2L, 1L, 2L, 1L, 1L, 2L, 3L, 1L, 2L, 1L, 1L, 3L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L), recurrence = c(1L, 1L, 1L, 1L, 0L, 1L, 0L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L), rfs = c(19.1, 15.33333333, 
49.16666667, 15.6, 57.16666667, 47.63333333, 54, 16.93333333, 
6.7, 102.1, 24.33666667, 127.7666667, 100.6333333, 25.96666667, 
1.233333333, 13.1, 72.16666667, 62, 97.23333333, 199.1, 24.73333333, 
60.46666667, 10.43333333, 31.76666667, 28.96666667, 56.43333333, 
9.533333333, 114.9333333, 114.8666667, 85.06666667, 107.6, 121.2, 
69.56666667, 70.03333333)), .Names = c("age", "WHO", "recurrence", 
"rfs"), class = "data.frame", row.names = 271:304) 

谢谢， 最好。

Answer 1

解决方案是

d$limits$WHO[2] <- 1
model <- cph(Surv(rfs,recurrence)~age + WHO,data=w)
summary(model)