Escaping special characters with etable.

[dlindzee]

Hi and thanks for such an amazing package. There seems to be an issue with escaping special characters in the etable function when the TEX option is set to true.

#Load the libraries
library(data.table)
library(fixest)

#Set the seed
set.seed(1)
 
X <- rep(c(1,1,2,3), 10) + rnorm(10*4,0,1)
a = rep(c("b","b","a", "c"), 10)
Y = rep(c(3,3,2,1.1), 10) + 0.5*X + rnorm(10*4,0,0.01)

#Create the data
DT = data.table(
  ID_var = a,
  Y,
  X
)

#Some regressions
Reg_1 <- feols(Y~X, DT)
Reg_2 <- feols(Y~X|ID_var, DT)

#Export the regressions
etable(Reg_1, Reg_2, tex = TRUE, fixef_sizes = T, se = "cluster", cluster = "ID_var")

I receive the following output.

\begin{table}[htbp]\centering
\caption{no title}
\begin{tabular}{lcc}
 & & \tabularnewline
\hline
\hline
Dependent Variable:&\multicolumn{2}{c}{Y}\\
Model:&(1)&(2)\\
\hline
\emph{Variables}\tabularnewline
(Intercept)&3.1110$^{***}$&  \\
  &(0.2996)&  \\
X&0.0468&0.5024$^{***}$\\
  &(0.0975)&(0.0023)\\
\hline
\emph{Fixed-Effects}&  & \\
ID\_var&No&Yes\\
\hline
\emph{Fit statistics}&  & \\
Observations& 40&40\\
# ID\_var&--&3\\
R$^2$ & 0.01063&0.99975\\
Within R$^2$ & --&0.99958\\
\hline
\hline
\multicolumn{3}{l}{\emph{One-way (ID_var) standard-errors in parentheses.}}\\
\multicolumn{3}{l}{\emph{Signif Codes: ***: 0.01, **: 0.05, *: 0.1}}\\
\end{tabular}
\end{table}

Notice that ID_var in the clustered error variable in the fourth last line of the table should be escaped as ID\_var. More importantly the # in the fit statistics section of the table, under the Observations variable, also not escaped. This bug persists even if one supplies a file instead of setting tex = TRUE.

Thanks again.

[lrberge]

Thanks @dlindzee !

I thought I had got rid of all escaping problems... but they are still lurking around the corner!
I've fixed this in the dev version. Hopefully we should get closer!

Thanks again for the report!