Training: File Naming and Coding Conventions

LAERI File Naming and Coding Conventions

Click through the module for an introduction to LAERI’s file and variable naming conventions and coding style. For questions, email LAERI staff at laeri@g.ucla.edu.

File Structures and Naming Conventions

How are project files structured?

Every LAERI project has a main project folder and the following subfolders:

DO: Syntax files

DTA: Data files

LOG: Log files

OUT: Output (e.g., tables and graphs)

SOURCE: Source data created for the project (e.g., a list of schools that implemented a specific program)

DOCUMENTATION: Documentation relevant to the project

OLD: Archived files

How do I name syntax files?

LAERI uses a standard naming convention for all projects: projectname_overarchingtask_step#_subtask.

If you edit your syntax file, save the file with the new version number using the following naming convention: projectname_overarchingtask_step#_subtask_V#. Then move the old version to the “old” folder.

How do I name data and log files?

All files produced by a syntax file should have the same file name as the syntax file.

The version numbers of data and log files should match the version number of the syntax file that produced them.

Move old versions of data and log files into the “old” folder. The server has plenty of storage space. There is no need to delete old files.

Variable Naming and Labeling Conventions

Variable Name and Labeling Conventions

How do I name variables? What prefixes do I use?

LAERI uses a standard variable naming convention for data in the archive: prefix + substantive name + end year (e.g., adfem17).

Whenever possible, variable names should not exceed 13 characters.

Example Variable Name Prefixes

ac=course data
act=transcript data
ad=demographic data
ab=behavior data
aa=assessment data
sese=elementary School Experience Survey data
sess=secondary School Experience Survey data
ns=National Student Clearinghouse data

To learn more about LAERI variable name prefixes, see the LAERI data archive overview stored on the I: drive.

How do I name recoded variables?

When you recode a variable, make a new version of the variable. Do not overwrite the existing variable.

Example Syntax

// Recoding "Don't Know" responses, i.e., 99,  to missing
recode sesehappy18 99=., into(sesehappy18V2)

//label the variable
label var sesehappy18V2   ///
"I am happy to be at this school (Don't Know=.);sv:sesehappy18"

//check coding
tab sesehappy18 sesehappy18V2,  m

Additional Resources

Stata Manual: Creating New Variables

IDRE Stata Learning Module: Creating and Recoding Variables

How do I label variables?

How do I label values?

Categorical variables should have clear value labels.

Try to keep the value labels short so that the entire label displays when you tab the variable.

Additional Resources

Stata Manual: Label Command

Stata Manual: Notes Command

IDRE Stata Learning Module: Labeling Data

Coding Style: Syntax Files

A program that is perfectly clear today is only clear because you just wrote it. Put it away for a few months, and it will most likely take you a while to figure out what it does and how it does it. If it takes you a while to figure it out, how long would it take someone else to figure it out?

Lester McCann

How do I begin a syntax file? How do I write a header?

File Header

All LAERI syntax files include a header than explains the purpose of the program.

What do I include after the header? What is a program setup section?

All LAERI syntax files include a “program setup” section below the file header.

Click on the command names on the right to read the Stata manual entries for the commands.

clear all: Drops all data, value labels, matrices, scalars, and stored results from memory

macro drop _all: Drops all macros from memory

set type double: Tells Stata to store numeric variables in double format (~16 digits of accuracy)

set maxvar 10000: Tells Stata to allow up to 10,000 variables

set more off: Tells Stata not to pause or display a -more- message

set linesize 120: Sets the line width, in characters, for both the screen and the log file

version 16: Tells Stata which version of the program you were using when you wrote the syntax file (maintains backward compatibility)

How do I organize syntax files?

Syntax files should be:

self-contained and able to run as stand-alone programs
logically organized so that the reader can easily follow the author’s thought process
formatted in a way that makes them easy to understand (consistent spacing and indentation)

How do I comment on syntax files?

Additional Resources

Harvard Strategic Data Project Coding Style Guide

IDRE: Clean-up, Comments and Code – Making it Maintainable