Open danielrojaslozano opened 1 year ago
Hello, I'm running the following code and it works
out_mc_riot = gsynth(vacancies_pc ~ treat_riot + random_x, data = dt, estimator = 'mc', index = c("codmun","time"), force = "two-way", CV = TRUE, r = c(0, 5), se = TRUE, cl = c('codmun','coddep_month'), inference = "nonparametric", parallel = FALSE)
Then, I subset dt to select covariates and omit NAs:
dt_covariates = dt[, .(vacancies_pc,random_x,treat_riot,log_urban_pop, codmun,coddep_month,time)] %>% na.omit()
Then, I run the model with log_urban_pop:
out_mc_riot2 = gsynth(vacancies_pc ~ treat_riot + random_x + log_urban_pop, data = dt_covariates, estimator = 'mc', index = c("codmun","time"), force = "two-way", CV = TRUE, r = c(0, 5), se = TRUE, cl = c('codmun','coddep_month'), inference = "nonparametric", parallel = FALSE)
And I get: Error in [.data.frame(data, , variable) : undefined columns selected
[.data.frame
Why is this happening? Note that I don't get the error message when estimator = 'ife' and inference = 'parametric'
Any ideas?
Hello, I'm running the following code and it works
out_mc_riot = gsynth(vacancies_pc ~ treat_riot + random_x, data = dt, estimator = 'mc', index = c("codmun","time"), force = "two-way", CV = TRUE, r = c(0, 5), se = TRUE, cl = c('codmun','coddep_month'), inference = "nonparametric", parallel = FALSE)
Then, I subset dt to select covariates and omit NAs:
dt_covariates = dt[, .(vacancies_pc,random_x,treat_riot,log_urban_pop, codmun,coddep_month,time)] %>% na.omit()
Then, I run the model with log_urban_pop:
out_mc_riot2 = gsynth(vacancies_pc ~ treat_riot + random_x + log_urban_pop, data = dt_covariates, estimator = 'mc', index = c("codmun","time"), force = "two-way", CV = TRUE, r = c(0, 5), se = TRUE, cl = c('codmun','coddep_month'), inference = "nonparametric", parallel = FALSE)
And I get: Error in
[.data.frame
(data, , variable) : undefined columns selectedWhy is this happening? Note that I don't get the error message when estimator = 'ife' and inference = 'parametric'
Any ideas?