如何将 N、对数似然、AIC 和 BIC 值分配给合并的“gtsummary”table 输出中的每个多元回归模型？

Question

请问，如何将 N、对数似然、AIC 和 BIC 值分配给合并的 gtsummary table 输出中的每个多元回归模型？

我是一个 Rmerteur 试图拟合加性多变量模型，向后续模型添加更高级别的变量（参见下面的示例）。同时，一旦我合并了单独的模型，我就无法保留每个模型拟合的 N、对数似然、AIC 和 BIC 值。我需要估计值来比较模型之间的拟合度。

有没有简单的方法来完成这个？

感谢您的帮助！

m1<-
  tbl_regression(mod1,
    exponentiate = T, hide_n = T, 
    add_estimate_to_reference_rows = T,
    estimate_fun = ~style_ratio(.x, digits = 2), 
    pvalue_fun = ~style_pvalue(.x, digits = 2), 
    label = list(age_yr_grp ~ "Age",
                 educ ~ "Education")) %>%
  bold_p(t = 0.05) %>% 
  add_glance_source_note()  

m2<-
  tbl_regression(mod2,
    exponentiate = T, hide_n = T, 
    add_estimate_to_reference_rows = T,
    estimate_fun = ~style_ratio(.x, digits = 2), 
    pvalue_fun = ~style_pvalue(.x, digits = 2), 
    label = list(age_yr_grp ~ "Age",
                 educ ~ "Education",
                 hhwealth ~ "Household wealth",
                 hhsize ~ "Household size")) %>%
  bold_p(t = 0.05) %>% 
  add_glance_source_note()  

m3<-
  tbl_regression(mod3,
    exponentiate = T, hide_n = T, 
    add_estimate_to_reference_rows = T,
    estimate_fun = ~style_ratio(.x, digits = 2), 
    pvalue_fun = ~style_pvalue(.x, digits = 2), 
    label = list(age_yr_grp ~ "Age",
                 educ ~ "Education",
                 hhwealth ~ "Household wealth",
                 hhsize ~ "Household size",
                 placeres~ "Area of residence")) %>%
  bold_p(t = 0.05) %>% 
  add_glance_source_note()     


tab3_multivars_models= 
  tbl_merge(list(m1,m2,m3),
  tab_spanner= c("**Mod 1**","**Mod 2**","**Mod 3**"))%>%
  as_gt() %>%
  tab_header(title = "Table title here") %>%
  tab_source_note(gt::md("*Notes here"))

gt::gtsave(tab3_multivars_models, "dir\tab3_multivars_models.html", inline_css = TRUE)

Answer 1

您需要使用 add_glance_table() 而不是源注释版本。

library(gtsummary)
#> #Uighur
packageVersion("gtsummary")
#> [1] '1.4.0'

tbl1 <-
  lm(age ~ marker + grade, trial) %>% 
  tbl_regression() %>%
  add_glance_table(
    label = list(sigma ~ "\U03C3"),
    include = c(r.squared, AIC, sigma)
  )

tbl2 <- tbl_merge(list(tbl1, tbl1))

^{由 reprex package (v2.0.0)}

于 2021-04-17 创建

如何将 N、对数似然、AIC 和 BIC 值分配给合并的“gtsummary”table 输出中的每个多元回归模型？

How to assign the N, Log-likelihood, AIC, and BIC values to each of multivariate regression models in a merged `gtsummary` table output?

r

gtsummary