0

I am trying to do a tbl regression but i cannot add the global p and there is an error. It seems to be caused by one parameter "groupe N" because without it there is no error. I coded this parameter whit case_when.

Creation of  "groupe N" :

d <- d %>%
  mutate(groupeN = case_when(
    chir_ypN == 0  & `chir_Nb_N0 _cPR`== 0 & `chir_Nb_N+_PR` == 0 ~"ypN0", 
    chir_ypN == 0 & `chir_Nb_N0 _cPR`== 1 ~ "ypN0/RC",
    chir_ypN != 0 & `chir_Nb_N+_PR` == 0 & `chir_Nb_N0 _cPR`== 0 ~ "ypN+",
    chir_ypN != 0 ~  "ypN+/RP"
  ))

-----------------------
Multivariate analysis : 

PFS_cox <- coxph(
  Surv(time, relapse) ~ 
    diag_CPS_PDL1+
    Groupe_ypT+
    chir_ypN+
    chir_Embol_lymph+
    chir_Pn+
    groupeN,
  data = data_PFS_cox
  )

PFS_multi <- tbl_regression(PFS_cox, exponentiate = TRUE) %>%
  add_global_p(keep=TRUE)

PFS_combi <- tbl_merge(tbls= list(PFS_univ, PFS_multi),
                       tab_spanner = c("Univariate", "Multivariate"))

PFS_multi
PFS_combi


------------------------------------

The error message :

✖ There was an error running `car::Anova()`, likely due to this model type not being supported. The results displayed are based on `add_global_p(anova_fun = gtsummary::tidy_wald_test)`
✖ `add_global_p(anova_fun = gtsummary::tidy_wald_test)` failed...
Error in dplyr::mutate(., model_terms = unlist(.data$data[["model_terms"]]) %>%  : 
  ℹ In argument: `wald_test = `%>%`(...)`.
ℹ In row 6.
Caused by error in `qr.solve()`:
! singular matrix 'a' in solve
In addition: Warning message:
In stop(e, call. = FALSE) : additional arguments ignored in stop()

If you could help me figure out this problem i will be very gratefull ! Thank you.

Ldkr
  • 1

0 Answers0