How to calculate robust standard errors in stargazer for plm (pooling) model in R?

Question

I have several regression models using plm and pooling. My data is a pooled cross sectional / time series data, with data on bond issuances. The data consists of observations of around 2000 bond issuances, with around 25 bond-descriptive variables.

I want to calculate the robust standard errors for one or all of the regression models, in order to add it in my stargazer visualization. My regressions are as follows:

#Regression    
primaryreg4 <- plm(issueyield ~ issuer + exchange + yearmonth + maturity.cat + size.cat + coupontype, 
                   data = data, 
                   index = c("ID", "issuedate"), 
                   model = "pooling")

All the variables are fixed effects (FE), so they are dummy variables. I want to calculate the robust standard errors of this model, and add it to stargazer. The code I have tried in order to calculate the standard errors is:

cov.r4 <- vcovHC(primaryreg4, type = "HC3")
robust_se_r4 <- sqrt(diag(cov.r4))

# or (it is the same)

robust_se_r4 <- sqrt(diag(vcovHC(primaryreg4, type = "HC3")))

I would then specify the standard errors in stargazer to be robust_se_r. However, i get this error message:

Error: cannot allocate vector of size 15.8 Gb

Does anyone know how to solve this? I understand that it is a memory problem, but the file really should not be too big - my data is about 2000 observations of 25 variables (not too big!).

Can you make the data available somewhere? Or provide a self-contained reproducible example? — Helix123

Marco Marco · Accepted Answer · 2019-12-03T11:51:18

Your code is running on minimal panel data:

library("plm")
data("Produc", package = "plm")

# Regression    
model <- plm(log(gsp) ~ log(pcap) + log(pc) + log(emp) + unemp,
          data = Produc, 
          index = c("state","year"),
          method="pooling")

# Robust Standard Errors
robust_se <- sqrt(diag(vcovHC(model, type = "HC3")))

Can you describe your variable more in detail? What categories does maturity and size have? Try glimpse from dplyr to check the data type and format of each variable.

How to calculate robust standard errors in stargazer for plm (pooling) model in R?

1 Answers