Book

"A Practitioner's Guide to Stochastic Frontier Analysis Using Stata" (with Subal C. Kumbhakar and Alan Horncastle), Cambridge University Press, January 2015.

• Cambridge University Press's information page; you may also place orders here.
• Purchase from Amazon.
• 華通書坊， 新竹市大學路1001號，交通大學內活動中心二樓，Tel: 03-5720317, Email: huatungbook@gmail.com.

Stata programs for the Book

I wrote many of the Stata commands used in the book. You may install these command files (*.ado), demo files (*.do), and demo dataset (*.dta) by following instructions on this page which is on a Google site. If you cannot access the Google site (for whatever reason), here is a short instruction.

1. Copy and paste one of the following lines (in one line) to Stata's Command window and execute it.
   • To download the package from the Google server:
     net install sfbook_install, from(https://sites.google.com/site/sfbook2014/home/install/) replace
   • To download the package from the server at National Taiwan University:
     net install sfbook_install, from(http://homepage.ntu.edu.tw/~wangh/sfbook/) replace
2. Execute the following command in Stata.
   sfbook_install
   It will install the programs (*.ado files) to Stata's PLUS directory, and copy the demo files (*.do, *.dta) to c:\sfbook_demo for Windows and /users/`c(username)'/sfbook_demo for Mac.
3. After the programs are installed, you may use help sfbook to see a short description on the installed Stata commands.

---

SFrontiers.jl: A Julia Package for Estimating Stochastic Frontier Models

I also wrote a package for estimating stochastic frontier models in the computer language Julia.

1. What is Julia? A free and open source software built with scientific computing and data science in mind.
2. Why Julia? Speed and accuracy.

Compared to the Stata commands, SFrontiers.jl does not have the system models. On the onther hand, it covers all of the cross-sectional and panel data models, including some panel models that have not been added to the Stata suit. For all of the included models, it could compute marginal effects of inefficiency determinants (if any) and bootstrap the standard errors of the mean marginal effects.

Find out more details here! The webpage includes an example in Jupyter notebook for easier exploration.