Tutorials
In each tutorial, we will work through a real-world data journalism workflow — starting from raw source data and combining tables step by step to produce a consolidated table ready for analysis. The workflows are drawn from journalists at leading news organizations, including The New York Times, ProPublica, BuzzFeed News, and more. Each was originally carried out in a programming language; here we reproduce the same result using OpenRoundup’s click-based interface.
Types of workflows
We categorize workflows based on the type of operations they use to consolidate tables together:
- stack-only: Only uses stack operations to combine tables by appending rows together.
- pack-only: Only uses pack operations to combine tables by joining columns together.
- hybrid: Combines both stack and pack operations in the same workflow.
Getting started
These tutorials provide an easy-to-follow introduction to using OpenRoundup with real-world data and workflows.
U.S. Refugee Resettlement
workflow type: Pack-only
In this tutorial, we will combine a table of refugee arrivals by state with a table of U.S. Census population estimates using a pack operation. Along the way, we will encounter unmatched rows between the two tables and learn how to inspect and resolve those mismatches.
Crime and Heat
workflow type: Hybrid
In this tutorial, we will combine two NOAA weather tables and one crime incident table into a single result that links each crime record to the maximum temperature on that day. We will use a stack operation to merge the weather tables, then a pack to join the combined weather data with the crime data.
Demolition Permits in Austin
workflow type: Stack-only
In this tutorial, we will combine two construction permit tables — full and partial demolitions — into a single composite table using a stack operation. We will also add a derived column to each source table before stacking so the combined result distinguishes between demolition types.
Next steps
Once you’ve got the basics down, these tutorials will help you learn how to use OpenRoundup for more complex workflows.
Voter Registration
workflow type: Stack-only
In this tutorial, we will incrementally assemble 11 tables of annually published voter registration data by composing multiple stack operations together, producing a single consolidated table ready for analysis.
California H-2A Temporary Agricultural Worker Visas Analysis
workflow type: Stack-only
In this tutorial, we will stack nine years of H-2A agricultural worker visa data into a single consolidated table, resolving schema drift as we align columns across tables from different years.
Drug Payment Reports
workflow type: Stack-only
In this tutorial, we will combine twelve years of pharmaceutical payment records into two consolidated tables, first resolving a schema mismatch in an external tool before stacking the tables in OpenRoundup.
IRS Audit Rates by County
workflow type: Hybrid
In this tutorial, we will stack four years of county tax-filing data and then pack the result with a table of IRS audit rates by county — discovering and resolving a FIPS code mismatch between data sources along the way.
Democratic Candidate Co-Donors
workflow type: Hybrid
In this tutorial, we will stack 15 FEC campaign filing tables together, then pack the result twice — first with filing metadata and then with candidate metadata — to produce a consolidated table of co-donors.
Babyname Politics
In this tutorial, we will stack 51 state-level baby name tables into a single dataset of approximately 6.6 million rows, then pack the result with 2016 presidential election data. This tutorial is ideal for exploring OpenRoundup’s scalability with a large number of identically-schemed source tables.
workflow type: Hybrid
Last Updated: June 11, 2026