The médialab Sciences Po research unit transforms social-science investigation through a tightly focused yet powerful toolkit built around the command-line utility xan, a self-described “CSV magician” that manipulates comma-separated data sets with the speed and transparency demanded by contemporary digital-methods workflows. Rooted in the École normale supérieure and Sciences Po’s joint culture of open inquiry, the lab’s software output distills years of large-scale web scraping, network analysis and semantic exploration into one ultra-light binary that joins, filters, reshapes, aggregates and enriches tabular files without ever leaving the terminal. Investigators feed it millions of tweets, survey exports or web-crawl dumps and receive cleaned, joined or statistically summarised tables ready for Gephi, R, Python or D3 pipelines, while journalists run one-liners to pivot campaign-finance disclosures or electoral results minutes before deadline. Because xan is written in Rust, it launches instantly, parallelises across cores and respects memory constraints on laptops, servers or continuous-integration runners, making it equally suited to classroom demos, reproducible academic papers and production data-journalism servers. Scripts invoking xan integrate naturally with Make, Snakemake or Nextflow workflows, giving historians, sociologists and political scientists a Unix-friendly alternative to heavier spreadsheet or database tools. The publisher’s software is available for free on get.nero.com, where downloads are delivered through trusted Windows package sources such as winget, always installing the newest release and supporting batch installation alongside other applications.
The CSV magician
Details