CODEWITHSUNDEEP
rregressionlmmodelsummaryfixest
Dawit G
Dawit G

Reputation: 11

Using fixest/etable to stack regression results from feols to create stacked panels

How can I vertically stack two different sets of regression outputs using the same set of dependent variables from fixest/feols using etable? For instance, suppose you estimate the same specification with different samples

est1 <- feols(c(y1,y2,y3,y4) ~ x1, data %>% filter(sample=="A"))

est2 <- feols(c(y1,y2,y3,y4) ~ x1, data %>% filter(sample=="B"))

I would like to stack the results in est1 and est2 vertically (instead of horizontally) to create a latex table so that Panel A has the results from est 1 (using sample "A") and Panel B of the same table has the results from est2 (without having to repeat the dependent variable names).

I have tried to do this using "modelsummary" instead of etable. Unfortunately, for IV regression, I was not able to export all the relevant statistics to modelsummary from the fixest_multi object.

Upvotes: 1

Views: 589

Answers (1)

mikeytop
mikeytop

Reputation: 170

You can use the panelsummary package.

library(fixest)
library(tidyverse)
library(panelsummary)


est1 <- feols(c(cyl,hp,drat) ~ mpg, data= mtcars %>% filter(wt <= 2))

est2 <- feols(c(cyl,hp,drat) ~ mpg,data =  mtcars %>% filter(wt > 2))

panelsummary(list(est1),
             list(est2),
             panel_labels = c("Panel A: Restricting WT <= 2",
                              "Panel B: Restricting WT > 2"))

Full disclosure: I am the author of the panelsummary package.

Upvotes: 0

Related Questions