CODEWITHSUNDEEP
rsurvival-analysis
Andrew James
Andrew James

Reputation: 67

Can I prevent summary.coxph from truncating coefficient names

After running a cox proportional hazards model with penalized splines, the summary table truncates the row names of the coefficients. Is there a way to print the full parameter or to increase the truncation limit to ensure full row names.

Modeling some dummy data:

library(survival)

dummy = data.frame(
  survival = c(0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,1), 
  time = c(60,120,180,240,300,360,60,120,180,240,60,120,180,240,300,360,420,480,60,120,180,60,120,180,240), 
  cov1 = c(9.1,7.3,8.2,0.9,0.7,8.7,6.3,7.6,1.5,2.9,1.9,1.3,6.4,1.3,8.9,1.6,7.8,1.2,0.8,5.4,3.5,2.2,4.8,9.3,5.5), 
  covariate2 = c(323,378,261.5,324,492,246.5,330.5,268,77.5,254.5,216,224,206,159.5,76,227.5,444,379,71,232,139,461,366,380,350.5))

mod_survival <- coxph(Surv(time, survival) ~ pspline(cov1) + pspline(covariate2, df = 2), data = dummy)
summary <- summary(mod_survival)
summary$coefficients

Results in the following:

> summary$coefficients
                                 coef se(coef)      se2        Chisq         DF         p
pspline(cov1), linear      0.01186380 4.180302 4.042494 8.054382e-06  1.0000000 0.9977356
pspline(cov1), nonlin              NA       NA       NA 1.090602e-01  6.4931508 0.9999908
pspline(covariate2, df =  -0.04858999 8.095739 7.212111 3.602306e-05  1.0000000 0.9952112
pspline(covariate2, df =           NA       NA       NA 2.154301e-10 -0.2063786 0.9964457

How can I print out the something like below?

                                              coef se(coef)      se2        Chisq         DF         p
pspline(cov1), linear                   0.01186380 4.180302 4.042494 8.054382e-06  1.0000000 0.9977356
pspline(cov1), nonlin                           NA       NA       NA 1.090602e-01  6.4931508 0.9999908
pspline(covariate2, df = 2), linear    -0.04858999 8.095739 7.212111 3.602306e-05  1.0000000 0.9952112
pspline(covariate2, df = 2), nonlin             NA       NA       NA 2.154301e-10 -0.2063786 0.9964457

Upvotes: 2

Views: 84

Answers (1)

David L Thiessen
David L Thiessen

Reputation: 694

The maxlabel argument in the summary.coxph.penal() function controls the length of coefficient names. You can see the source code by running getAnywhere(summary.coxph.penal) after loading the survival package. The default value is maxlabel = 25, as you can see from the example.

Changing the summary() call to summary <- summary(mod_survival, maxlabel = 35), or a greater number, solves your issue.

Upvotes: 1

Related Questions