Skip to content

ndexr from the bottom up

This is the index of everything. R knowledge, infrastructure patterns, the story of ndexr — assembled from recordings made over years of building. Start anywhere. The chapters have an order but the story has no ending.

Setup

Every production system starts the same way — someone opens a terminal and types. These are those moments. R and RStudio, git, Docker, WSL. Nothing clever here, but nothing skippable either. The difference between a data scientist and an engineer is usually just this: one of them has done this part enough times that it's boring.

R in Production

The R in Production series was an attempt to write down the gap between R as a data science tool and R as something you'd trust with real work. Packages, version control, PostgreSQL, Python interop — six episodes that treat R the way it deserves to be treated. The discipline here is what makes everything else in ndexr possible.

Infrastructure

Infrastructure is where most R projects go to die. Not because it's hard, but because no one writes it down. These recordings cover the actual work: spinning up EC2, configuring HTTPS, running NGINX in front of Shiny, Docker Compose on production iron, RStudio on AWS. The parts that are boring until they break — and then they're everything.

Building ndexr

These are the live sessions. Unedited, unglamorous, watching the platform actually take shape. A better Shiny template. R infrastructure managed from a phone. An HPC spun up in under ten minutes. This is what building looks like — not the polished version, the real one. The decisions, the restarts, the parts that worked on the first try and the parts that didn't.