Sleep Medicine Epidemiology
Brigham and Women's Hospital · Harvard Medical School

Sleep Portal 0.15.0

Spout JSON Data Dictionaries


The 0.15.0 release has simplified the data dictionary specification process by integrating with Spout: Tested JSON Data Dictionaries.

Leveraging Spout for the SHHS Data Dictionary

Spout is a Ruby Gem that allows CSV data dictionaries to be converted into a corresponding JSON git repository. Spout provides a series of tests to assure variable and domain name uniqueness, variable type validation, and also tests to assure proper JSON formatting.

By breaking down a CSV into multiple JSON files (one for each variable and one for each domain), we are given the freedom to allow better collaboration on a data dictionary, while maintaining the integrity of the data dictionary itself.

For the Sleep Portal, our primary data dictionary is the SHHS Data Dictionary. Through the use of Spout and GitHub Issues, we were able to discover and address 46 issues across 124 commits for our v0.1.0 release of the SHHS Data Dictionary that is used in the current release of the Sleep Portal.

We were also able to discover data collection nuances in the SHHS overall variable in Issue #40.SHHS Data Dictionary Issue #40 - Sleep Portal

Spout has allowed us to greatly simplify the internals in the Sleep Portal, contributing to a reduction in over 2,400 lines of code from the previous 0.14.2 release.

Additionally, we have also cleaned up the Sleep Portal to provide a cleaner interface to allow the researcher to focus on searching and visualizing data in the SHHS dataset.

A list of changes includes:

0.15.0 (July 26, 2013)



Bug Fix