I ran an binomial logit model while controlling for bank- and time-fixed effects by using the feglm()
function from the fixest package. However this function saves the regression output as a list. Unfortunately Stargazer doesn't seem to recognize this. Can anyone help me with this?
This is the regression I ran:
BinomialLogitModelWinsorizedLag <- feglm(CoCoIssuance ~ log(TotalAssetsLagged) + DepositsOverAssetsLagged + T1RatioLagged + AssetGrowthLagged + EquityOverAssetsLagged + CashOverAssetsLagged + CET1RatioLagged + DividendsOverEquityLagged + PriceOverBookLagged + ROALagged + LoansOverAssetsLagged | IssuerName + YearOfIssuance, data = WinsorizedDataLagged , family = binomial("logit") )
I tried extracting the estimated coefficients and standard errors from this list and save it as a data frame. However when I do this, Stargazer gives me the descriptive statistics and not the regression output table that I want.
# extract coefficients and standard errors
coef_tableLogitLag <- BinomialLogitModelWinsorizedLag$coeftable
stargazer(coef_tableLogitLag, type = "text")
I am not sure about stargazer, why it is not producing the result what you wanted, I would suggest you to try out texreg
package instead.
Below code work fine for me:
library(fixest)
library(texreg)
# Fit a binomial logit model with bank- and time-fixed effects
model1 <- feglm(am ~ mpg + cyl + hp | factor(vs) + factor(gear),
data = mtcars, family = binomial("logit"))
htmlreg(model1, file = "texreg.html",
inline.css = FALSE, doctype = TRUE, html.tag = TRUE,
head.tag = TRUE, body.tag = TRUE)
This produces the result as:
Or for the models:
# Fit a binomial logit model
model1 <- feglm(am ~ mpg + cyl + hp | factor(vs) + factor(gear),
data = mtcars, family = binomial("logit"))
model2 <- feglm(am ~ mpg + cyl + hp + factor(vs) + factor(gear),
data = mtcars, family = binomial("logit"))
The code below:
htmlreg(list(model1, model2), file = "texreg2.html",
inline.css = FALSE, doctype = TRUE, html.tag = TRUE,
head.tag = TRUE, body.tag = TRUE)
Produces: