losing the variable name after pmap?

[von_olaf]

Consider this simple example

get_reg <- function(...) {
  myformula <- paste(..1, ..2)
  lm(myformula, data = iris)
}

formulas <- c('~ Sepal.Length ', '~Sepal.Length + Sepal.Width')

list(depvar = 'Petal.Width', 
     myformula = formulas) %>% 
  cross_df() %>% 
  pmap(., ~lm(get_reg(..1, ..2))) %>% 
  stargazer(type = 'text')

which gives:

=====================================================================
                                   Dependent variable:               
                    -------------------------------------------------
                                           ..1                       
                              (1)                      (2)           
---------------------------------------------------------------------
Sepal.Length                0.753***                 0.723***        
                            (0.044)                  (0.039)         
                                                                     
Sepal.Width                                         -0.479***        
                                                     (0.074)         
                                                                     
Constant                   -3.200***                -1.563***        
                            (0.257)                  (0.339)         
                                                                     
---------------------------------------------------------------------
Observations                  150                      150           
R2                           0.669                    0.743          
Adjusted R2                  0.667                    0.739          
Residual Std. Error     0.440 (df = 148)         0.389 (df = 147)    
F Statistic         299.167*** (df = 1; 148) 212.412*** (df = 2; 147)
=====================================================================
Note:                                     *p<0.1; **p<0.05; ***p<0.01

I do not understand why I am losing the dependent variable name. As you can see in the regression table, the dependent variable is the cryptic ..1 instead of Petal.Width.

Any ideas? Thanks!

[von_olaf]

The wrapper I wrote actually works. I just have to call it correctly.

list(depvar = 'Petal.Width', 
     myformula = formulas) %>% 
  cross_df() %>% 
  pmap(., ~get_reg(..1, ..2)) %>% 
  stargazer(type = 'text')

The interesting takeaway is that the formula needs to be computed outside of lm, otherwise the dependent variable label gets lost.