DEA for 9040 firms

[Denila Jinny]

Hi all, I am trying to do DEA (Data Envelopment Analysis) for 9040 firms with 5 inputs and 2 outputs. STATA seems to hang when I run this model after matsize is set to maximum. Is there any other quick and efficient way to do this? I just need the efficiency score (or relative efficiency) out of the analysis.

[Phil Bromiley]

You'll increase your chances of a helpful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stata output, and sample data using dataex.

I have not used the DEA routine. We do ask that anyone using a user written routine identify the routine and where the individual found the routine. From your brief description, it is not clear whether there is a real problem or the routine is just taking a long time. When it hangs, is Stata using a lot of CPU? Are there other user-written Stata DEA routines that might do the job? I'd try the routine on a small subsample to make sure the number of firms is the problem. There are some ways to get rough ideas of estimation difficulty on DEA. You might try varying the sample size to see how solution time increases with sample size. Sometimes "hanging" just means you let it run overnight or over the weekend.

[Denila Jinny]

To anyone coming across this question, I used rDEA in R to get efficiency scores. It just needed 2 mins to give me the output.