Extract output from a list of stochastic frontier models and store in a single data frame-CodePudding

I have a list sfa_out with 8 stochastic frontier models stored inside.

I want to extract the efficiency estimate from each model using efficiencies() function and store the extracted efficiency measure for all models in a single data frame. Doing so individually for each model is not a problem, however, I want to write a function that does it for all models.

My attempt using a for loop is below:

# define an empty data frame
eff_out <- data.frame()

# write a for loop for each i model in the list "sfa_out"
for(i in 1:length(sfa_out$models)) {
  eff_out$i = as.data.frame(efficiencies(sfa_out$models[[i]])) %>%

# the code below pivots the data frame so that three columns are "col", "year" and "efficiency"
  
    mutate(col = row.names(efficiencies(sfa_out$models[[i]]))) %>%
    pivot_longer(cols = 1:23,
                 names_to = "year",
                 values_to = "efficiency") %>%
    drop_na()
}

However, doing so gives me the following error:

Error in `$<-.data.frame`(`*tmp*`, "i", value = list(col = c("GB0000365774",  : 
  replacement has 139 rows, data has 0

Any help here would be appreciated. Thanks.

CodePudding user response：

The following did the trick for me:

# create a new list by extracting only models from the list sfa_out
newlist <- as.list(sfa_out$models)

# define a function which gets the required output from each model

outputs <- function(newlist) {
  
# create an empty data.frame
  eff_out <- data.frame()
  
# loop through each model in the newlist
  for(i in 1:length(newlist)) {

      eff <- as.data.frame(efficiencies(newlist[[i]])) %>%
      mutate(col = row.names(efficiencies(newlist[[i]]))) %>%
      pivot_longer(cols = 1:23,
                   names_to = "year",
                   values_to = "efficiency") %>%
      drop_na()

# current model output  
   model_output <- data.frame(eff)
  
# append output to empty dataset we created before
   eff_out <- rbind(eff_out, model_output)
  }
  return(eff_out)
}
# apply the function to the list containing models
sfa_eff_outputs <- outputs(newlist)