High-Frequency Checks exercise, DIL Welcome Week 2026

MOCK, teaching version of an HFC pipeline, used in the Data Quality session (Thursday, July 23). It runs on fake data built to match the Household Water Questionnaire – V1 (see documentation/), the same case study used in the Data Ingestion, Cleaning & Tidying session.

main.do sets globals and runs one numbered do-file per check from code/. Each check exports its flags to one sheet of output/YYYY-MM-DD_HFC_Household_Water.xlsx — the report a field team would actually read.

Files 1 (import), 2 (duplicates) and 7 (dashboard) are completed, as a model of the style we use. Files 3–6 contain "YOUR TURN" stubs for you to complete during the exercise, using the DQAP your table filled in.

How to run it (option A, the GitHub workflow, recommended)

1. Install GitHub Desktop and sign in.
2. Clone this repository: green Code button → Open with GitHub Desktop.
3. In GitHub Desktop, create a new branch named after your table (e.g. table-3 or yourname-hfc). Work there, never on main.
4. Open main.do in Stata, add your username and file path to Section 2 (type di c(username) in Stata to see your machine's name), and run it end to end once as-is.
5. Complete do-files code/3-missingness.do to code/6-surveylogic.do (look for the YOUR TURN comments), re-running main.do as you go.
6. When everything runs, commit your changes in GitHub Desktop (do-files + today's report in output/), push your branch, and open a pull request to main so the facilitators can review your code. Full workflow details: see CONTRIBUTING.md.

How to run it (option B, no git, plain folder)

1. On the repository page, green Code button → Download ZIP.
2. Extract it anywhere on your computer.
3. In Stata: cd into the extracted folder, then do main.do. (With no username match, main.do falls back to the current working directory, no path edits needed.)
4. Complete do-files 3–6 the same way. You just won't be able to push; flag a facilitator at the end and we'll look at your code together.

Folder structure

main.do              runs everything (edit Section 2 paths first)
code/                one numbered do-file per check (+ ado/ for packages)
data/raw/            household_water_fake_raw.csv (fake!), _use.dta is local
documentation/       the questionnaire + the formatted report template
output/              today's excel report lands here

Requirements

• Stata 15 or newer. User-written commands (ieboilstart) are shipped in code/ado/ and wired up by main.do, nothing to install.

Development Innovation Lab · University of Chicago