CODEWITHSUNDEEP
rlatexregression
user1172558
user1172558

Reputation: 655

Manually create Latex Regression Table from R

I have a regression table that I want to export to Latex. But the object I'm working with does not support any traditional package like 'stargazer', or 'texreg'. I have to build it manually. This are the results:

                   Estimate   Std. Error    z value
(Intercept):1 -4.652887e-01 3.149742e-01  -1.477228
x1            -3.277175e-05 1.014935e-05  -3.228952
x2            -2.446376e-05 6.717509e-06  -3.641790
x3             1.666235e+00 5.228731e-01   3.186691
x4             2.507294e-08 5.276769e-09   4.751570

What I want is:

(Intercept)   -4.652887e-01 
              (3.149742e-01)
x1            -3.277175e-05 
              (1.014935e-05)
x2            -2.446376e-05 
              (6.717509e-06)
x3             1.666235e+00
              (5.228731e-01)
x4             2.507294e-08
              (5.276769e-09)

I could do this manually (see below), but I was looking for something more dynamic that I could apply in several contexts. 

res <- matrix(NA, nrow=2*dim(ctable), ncol=1)

res[1,] <- ctable[1,1]
res[3,] <- ctable[2,1]
res[5,] <- ctable[3,1]
res[7,] <- ctable[4,1]
res[9,] <- ctable[5,1]

res[2,] <- ctable[1,2]
res[4,] <- ctable[2,2]
res[6,] <- ctable[3,2]
res[8,] <- ctable[4,2]
res[10,] <- ctable[5,2]

Thanks!

Upvotes: 1

Views: 1478

Answers (1)

Djongs
Djongs

Reputation: 48

I dont know if I understood well your problem. But if you want just create your table dynamically, you can use the function below.

create_table = function(table_)
{
    p_table = data.frame(Variables=NA, Values=NA)
    for(i in 1:dim(table_)[1])
    {
        est_error = table_[i,1:2] # Im assuming that you know that 2 first columns are the values that you want
        name = c(rownames(table_)[i], '') # The final table rownames
        name_values = cbind(Variables=name, Values=est_error)
        p_table = rbind(p_table, name_values)
    }
    rownames(p_table) = 1:dim(p_table)[1]
    return(na.omit(p_table))
}

And you can use this function in the following way. Below, one example using the mtcars dataset.

ctable = summary(lm(mpg~. , data = mtcars))$coefficients
output_table = create_table(ctable)

The output is something like you posted. Then, is easy create a tex table from a R dataframe using the xtable package.

xtable::xtable(output_table)

Upvotes: 3

Related Questions