Skip to contents

Tidy a multi-state survival model

Source: R/custom_tidiers.R
tidy_coxphms.Rd

[Experimental] A tidier for multi-state models (survival::coxphms.object) generated with survival::coxph(). Term names will be updated to be consistent with generic models. The original term names are preserved in an "original_term" column. An additional column "state" will be added to provide status detail (i.e. the right part of original terms) and a column "y.level" will be populated by using values stored in x$states.

Usage

tidy_coxphms(x, conf.int = TRUE, conf.level = 0.95, ...)

Arguments

x

(coxphms)
A survival::coxphms.object model.

conf.int

(logical)
Whether or not to include a confidence interval in the tidied output.

conf.level

(numeric)
The confidence level to use for the confidence interval (between 0 ans 1).

...

Additional parameters passed to parameters::model_parameters().

See also

Other custom_tidiers: tidy_broom(), tidy_multgee(), tidy_parameters(), tidy_svy_vglm(), tidy_vgam(), tidy_with_broom_or_parameters(), tidy_zeroinfl()

Examples

# \donttest{
  library(survival)
  # dataset with competing-risk-style status
  df <- MASS::Melanoma
  df$id <-
    df |>
    row.names()
  df$sex <-
    df$sex |>
    factor(0:1, c("male", "female"))
  df$status <-
    df$status |>
    factor(c(2, 1, 3), c("alive", "died from melanoma", "dead from other causes"))

  mstate_model <- coxph(Surv(time, status) ~ sex, data = df, id = id)
  mstate_model |> tidy_coxphms()
#>    estimate std.error conf.level   conf.low conf.high statistic df.error
#> 1 0.6621781 0.2636234       0.95  0.1423868  1.181969  2.511833      203
#> 2 0.6302185 0.5283391       0.95 -0.4115176  1.671955  1.192830      203
#>     p.value original_term      term state                y.level
#> 1 0.1129944 sexfemale_1:2 sexfemale   1:2     died from melanoma
#> 2 0.2747595 sexfemale_1:3 sexfemale   1:3 dead from other causes
  mstate_model |> tidy_plus_plus()
#> ℹ <coxphms> model detected.
#> ✔ `tidy_coxphms()` used instead.
#> ℹ Add `tidy_fun = broom.helpers::tidy_coxphms` to quiet these messages.
#> # A tibble: 4 × 26
#>   group_by     y.level term  original_term variable var_label var_class var_type
#>   <fct>        <chr>   <chr> <chr>         <chr>    <chr>     <chr>     <chr>   
#> 1 died from m… died f… sexm… NA            sex      sex       factor    dichoto…
#> 2 died from m… died f… sexf… sexfemale_1:2 sex      sex       factor    dichoto…
#> 3 dead from o… dead f… sexm… NA            sex      sex       factor    dichoto…
#> 4 dead from o… dead f… sexf… sexfemale_1:3 sex      sex       factor    dichoto…
#> # ℹ 18 more variables: var_nlevels <int>, contrasts <chr>,
#> #   contrasts_type <chr>, reference_row <lgl>, label <chr>, n_obs <dbl>,
#> #   n_ind <dbl>, n_event <dbl>, exposure <dbl>, estimate <dbl>,
#> #   std.error <dbl>, conf.level <dbl>, conf.low <dbl>, conf.high <dbl>,
#> #   statistic <dbl>, df.error <int>, p.value <dbl>, state <chr>
# }