Improving pairwise_survdiff() to handle a formula with multiple variables

[kassambara]

No description provided.

[kassambara]

Now, pairwise_survdiff() supports a formula with multiple variables.

library(survival)
library(survminer)
pairwise_survdiff(Surv(time,status) ~rx+adhere, data = colon)

	Pairwise comparisons using Log-Rank test 

data:  colon and rx + adhere 

                     rx=Obs, adhere=0 rx=Obs, adhere=1 rx=Lev, adhere=0 rx=Lev, adhere=1
rx=Obs, adhere=1     0.100            -                -                -               
rx=Lev, adhere=0     0.791            0.100            -                -               
rx=Lev, adhere=1     0.144            0.791            0.142            -               
rx=Lev+5FU, adhere=0 4.5e-06          4.5e-06          1.6e-05          4.5e-06         
rx=Lev+5FU, adhere=1 0.791            0.157            0.791            0.192           
                     rx=Lev+5FU, adhere=0
rx=Obs, adhere=1     -                   
rx=Lev, adhere=0     -                   
rx=Lev, adhere=1     -                   
rx=Lev+5FU, adhere=0 -                   
rx=Lev+5FU, adhere=1 0.083               

P value adjustment method: BH