I want to have my fixed effects capitalized in the modelsummary output. While I know I could rename the columns before estimating the regression, I wanted to know if there is an easier way to do this within the modelsummary::modelsummary function. I have tried the coef_map argument, but it doesn't seem to affect the table.

library(tidyverse)
library(modelsummary)
library(fixest)

mtcars %>% 
  feols(mpg ~hp | gear, data = .) %>% 
  modelsummary(stars = T, coef_map = c("hp" = "Horsepower",
                                       "mpg" = 'MPG',
                                       "gear" = "Gear"))

   

As shown in the example, I want the "gear" fixed effect to be capitalized without having to rename in my data (if possible).

CodePudding user response：

A modelsummary table is essentially composed of two parts. The top contains estimates and is controlled by the coef_* arguments. The bottom contains “goodness-of-fit” and other model characteristics, and is controlled by the gof_* arguments.

You can rename and reformat and reorder the GOF info with the gof_map argument. In this example, I use the tribble function to create a data.frame with the information needed to rename and capitalize. You can also use a list, or a plain data.frame. Please refer to the documentation.

library(tidyverse)
library(modelsummary)
library(fixest)

gm <- tribble(~raw, ~clean, ~fmt,
              "FE: gear", "FE: Gear", ~fmt)
mtcars %>% 
  feols(mpg ~hp | gear, data = .) %>% 
  modelsummary(output = "markdown",
               stars = T, coef_map = c("hp" = "Horsepower",
                                       "mpg" = 'MPG',
                                       "gear" = "Gear"),
               gof_map = gm)

Note: ^^ p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001