Workshop
R for Data Science: Foundations of Analytics
Rapid analytics and visualization for mission support
- Duration
- 5 days
- Format
- Live virtual or on-site
- Hours
- 9:30 AM – 5:30 PM ET
- Programming
- Beginner-friendly — no prior coding required
You will be able to
- Import, clean, join and validate mission-relevant data in R and RStudio
- Use the tidyverse to summarise, reshape and visualise quickly
- Build charts and maps that hold up in a briefing
- Produce repeatable, parameterised R Markdown and Quarto reports
- Run and interpret lightweight models — trend, regression, simple classification
- Work reproducibly, with OPSEC and handoff to a teammate in mind
Who it is for
- US and NATO personnel working in analysis, logistics and readiness
- Analysts who need to produce defensible numbers quickly
- Anyone currently doing this work in a spreadsheet and hitting its limits
What you need first
- Familiarity with defense or government operations is helpful
- No prior coding required
Agenda
Day 1
Foundations & setup
- Orientation (30m) — what decision support with R means in practice, and what success looks like for your role.
- R and RStudio tour (60m) — console, scripts, projects, file hygiene.
- Data in, clean, and ready (2h) — CSV, Excel and Parquet; fixing headers with janitor; dates with lubridate; missingness.
- Tidyverse basics I (90m) — dplyr verbs, grouping and summarising, and the quality checks that catch bad data early.
- Mini-lab (30m) — clean a logistics dataset and produce a one-page summary.
- Outcome: a clean R project, a reproducible script, and your first visuals.
Day 2
Wrangling deep dive & visualization
- Tidyverse basics II (90m) — joins, keys, cross-checks; rowwise versus vectorised, and the pitfalls of each.
- Reshape and compare (60m) — pivot_longer and pivot_wider; tidy tables for side-by-side readiness comparison.
- Visual basics with ggplot2 (90m) — lines, bars, densities, faceting, and labelling that survives a projector.
- Design for decision makers (60m) — clarity against complexity, encoding choices, and why chartjunk costs you credibility.
- Lab (30m) — turn your wrangled tables into two or three briefing-quality charts.
- Outcome: a tidy pipeline and a small chart pack fit for leadership slides.
Day 3
Time, trend & geospatial
- Time and trend (75m) — time-based summaries, rolling means, simple change detection, and communicating uncertainty honestly.
- Geospatial foundations (90m) — sf basics, CRS and projections, simple choropleths, routes and areas.
- Operational maps (60m) — tmap against leaflet; when static beats interactive.
- Reporting I (45m) — R Markdown and Quarto, chunk options, inline values, kable tables.
- Mini-exercise (30m) — add a time-series panel and a map to your report.
- Outcome: a report blending trend and map views that rebuilds itself on new data.
Day 4
Reporting, parameters & modeling lite
- Reporting II (75m) — parameters, child documents, reusable templates, HTML and PDF variants.
- Reproducibility and handoff (45m) — project structure, relative paths, minimal versioning.
- Modeling basics (75m) — linear and logistic regression, feature hygiene, and when simple genuinely beats complex.
- Fit, check, explain (60m) — broom tidiers; confidence against prediction intervals; interpretation that survives a question.
- Scenario workshop (45m) — a parameterised brief that toggles between units, regions and time windows.
- Outcome: a parameterised report and an interpretable risk-or-trend model you can explain on one slide.
Day 5
Capstone build & brief
- Predict then plan lab (60m) — predict a delay risk, translate it into an action option, then run sensitivity checks on it.
- Capstone build (2h) — team project end to end: data, wrangling, visuals including a map, a small model, a parameterised report.
- Brief and peer review (60m) — five-minute briefs, with peers stress-testing your assumptions and your clarity.
- Next steps (30m) — packaging the work, running a small pilot, sustaining it.
- Outcome: a decision-support product ready for leadership or exercise use.
Materials provided
- Slides, step-by-step notebooks and sanitized datasets (CSV and Parquet)
- Cheat sheets for tidyverse, ggplot2 and a geospatial quickstart
- Report templates in R Markdown and Quarto, with a short style guide
Security & classification
All materials are unclassified, and the scenarios and datasets are sanitized for training.
Request dates, or a private cohort.
This runs as a scheduled cohort or privately for a single organisation, virtual or on-site. Tell us the room — how many people, how senior, and what they need to be able to do afterwards — and we will tell you honestly whether this is the right course for them.