NetCourse 151

Introduction to STATA Programming

Content:

An introduction to Stata programming dealing with what most statistical software users mean by programming, namely, the careful performance of reproducible analyses.

Through a combination of lectures, example applications, and carefully chosen problems, the course addresses the full range of methods and techniques necessary to be most productive in the Stata environment.

Course Lecturer:

Bill Gould, President of StataCorp and Head of Development.

Course Leaders:

Kerry Kammire, technical services analyst at StataCorp
Jennifer Rolfes, technical services representative at StataCorp

Course Dates:

March 19 - April 30, 2010

Cost:

£ 86 + VAT

Prerequisites:

  • Stata 11, installed and working
  • Basic knowledge of using Stata interactively
  • Internet web browser, such as Netscape, Microsoft Internet Explorer, or Mozilla, installed and working (course is platform independent)

Agenda

Lecture 1: Organization of analysis

  • Mechanics of using do-files, ado-files, and program define
  • Programming as automating data-management and analysis
  • Organization of multiple do-files
  • Ensuring reproducibility
  • Data verification (assert)
  • Debugging (trace)
  • Working with datasets too large to fit in memory
  • Indexing (subscripting) observations
  • assert as an alternative to branching
  • Reading and processing relational datasets
  • Consuming calculated results

Lecture 2: Macros, arguments, and looping

  • Using macros to contain strings
  • Using macros to contain numbers
  • Using macros to pass and receive arguments with do-files
  • Using macros to pass and receive arguments with programs
  • Distinction between local and global macros
  • Branching and looping
  • Program style
  • Shifting over lists
  • Looping across observations
  • Application: Listing data neatly (report writing and use of display)

There is an additional one-week break between Lectures 2 and 3 in this course because we have found the extra time is necessary for discussion.

Lecture 3: Examples and applications

  • Do-files, ado-files, and programs: when to use which
  • Complicated data management example: processing hospital admission/discharge data and merging with patient file
  • Working with dates
  • Working with dates and time of day
  • Programming bootstrapped standard errors
  • Programming Monte Carlo simulations
  • Use of postfile to program bootstraps and simulations beyond the capabilities of bstrap and simulate

Lecture 4: Ado-files

  • Organizing files and directories
  • Writing real ado-files: command to run analysis do-files
  • Automatically logging analysis
  • Putting time-and-date stamps on analyses
  • Error handling
  • Writing on-line help files
  • Temporary variables
  • Temporarily destroying data
  • Temporary files
  • Analysis-specific ado-files
  • General-purpose ado-files
  • Converting analysis-specific to general-purpose ado-files
  • More on display and producing formatted output
  • Parsing commands with syntax
  • Version control
©Timberlake Consultants Limited
Last revised:02/02/2010