36-613 · Data Visualization

Meetings: August 24–October 9, 2026 · Mondays and Wednesdays · 9:30–10:50 a.m. · Baker Hall 140A
Mini-course length: 7 weeks, 2 sessions per week
Teaching data: NOAA Storm Events, with optional HURDAT2 and enrichment data
Final-project data: global data, AI, and machine-learning salary archive, 2020-2025
Final product: client-ready workforce-compensation and recruiting report

Two-dataset structure: NOAA stays in the lectures and checkpoints. The final project uses a separate compensation archive for an executive-search client.

Course Goal

This course is about making visual analyses that help someone decide what to do. We will use R, ggplot2, and Quarto to audit public data, choose graphics that fit the question, support visual claims with statistical evidence, and explain results to a non-technical client. NOAA Storm Events supplies our shared classroom examples; the final report uses the separate workforce-compensation archive.

Recurring close read: Graphics critiques in the wild features three to five instructor-selected real-world graphics each week. Every pick credits and links to its original publication; the page does not republish paywalled graphics.

Week 0 · Setup and Starter Data

Complete before the first meeting.

Materials

Lab 00 · RStudio Setup

Topics

• R, RStudio, Quarto/R Markdown.
• Project folders and reproducibility.
• Loading NOAA Storm Events.
• What one row means.

You should finish with a rendered document that reports the data dimensions, lists event types, and includes one plot.

Week 1 · Aug 24–28 · Foundations and Categorical Data

Meetings

• Mon Aug 24 · remote · Why visualize? Tidy data and the grammar of graphics · source
• Wed Aug 26 · asynchronous, recorded · Principles and 1D categorical data · source

Topics

• Visualization as client communication.
• Tidy data and the grammar of graphics with ggplot2.
• Nominal versus ordinal categories; factor level order.
• Counts, proportions, and sampling uncertainty.

Homework

HW1 opens Sun Aug 23 · homework-01.qmd · PDF

Week 2 · Aug 31–Sep 4 · Categories to Distributions

Meetings

• Mon Aug 31 · Two categorical variables and one quantitative variable · source
• Wed Sep 2 · Quantitative distributions and density · source

Topics

• Two-way tables, mosaic plots, and heatmaps.
• Histograms, boxplots, ECDFs, and density estimates.
• KS tests, violin, and ridge plots.

Homework

HW1 due Tue Sep 1 · HW2 opens Sun Aug 30 · homework-02.qmd · PDF

Week 3 · Sep 7–11 · Two Quantitative Variables

Meetings

• Mon Sep 7 · no class, Labor Day
• Wed Sep 9 · Trends, residuals, and overplotting · source

Topics

• Regression, fitted trends, and residuals.
• Two-dimensional density, contours, and hexbins.

Homework

HW2 due Tue Sep 8 · HW3 opens Sun Sep 6 · homework-03.qmd · PDF

Week 4 · Sep 14–18 · High-Dimensional Data

Meetings

• Mon Sep 14 · Seeing many variables · source
• Wed Sep 16 · Distance, MDS, and PCA · source

Topics

• Pairs plots, correlations, and heatmaps.
• Distance, multidimensional scaling, and principal components.

Homework

HW3 due Tue Sep 15 · HW4 opens Sun Sep 13 · homework-04.qmd · PDF

Week 5 · Sep 21–25 · Embeddings, Time, and Space

Meetings

• Mon Sep 21 · Nonlinear embeddings and trends · source
• Wed Sep 23 · Time-series structure and spatial foundations · source

Topics

• t-SNE interpretation and stability.
• Daily and monthly seasonality; lags, smoothing, and autocorrelation.
• Point versus areal spatial data and projection.

Homework and assessment

HW4 due Tue Sep 22. Occasional brief checkpoints; no additional full homework from this point on.

Week 6 · Sep 28–Oct 2 · Maps, Design, and Text

Meetings

• Mon Sep 28 · Areal maps and graphic design · source
• Wed Sep 30 · Text as data · source

Topics

• Choropleths, normalization, and map joins.
• Event narratives as text data; word frequency, TF-IDF, or sentiment when useful.
• Titles that state findings; annotation, accessibility, color, and chart critique.

Assessment

Occasional brief in-class checkpoints.

Week 7 · Oct 5–9 · Animation, Critique, and Client Communication

Meetings

• Mon Oct 5 · Animation and interaction · source
• Wed Oct 7 · topic to be announced

Focus

We will compare static, animated, and interactive displays, practice evidence-based critique, and work on the recommendation, evidence, caveats, and next action in the final client report.

See final-project.md and project-rubric.md.

Assignments

All homework times are Eastern. Instructions and course materials are released on the teaching website. Open each Gradescope assignment from Canvas; submit homework as PDFs and the final project as one self-contained HTML file. Canvas contains official announcements, summary grades, solutions, and other restricted materials; Piazza hosts course questions and the two visualization critiques.

Final Project Requirements

The workforce-compensation report must include:

• 6-10 graphics.
• At least three client questions.
• At least three foundational graphics from the first half of the course.
• At least three advanced graphics from the second half of the course.
• No more than two one-variable plots.
• No more than three plots of the same type.
• At least one formal statistical analysis, model summary, or uncertainty display.
• A target employer and recruiting playbook.
• A visible audit of duplicates, coverage, provenance, and unsupported market claims.
• Clear recommendations, caveats, and future evidence requests.

How this course differs from 36-614

The 36-613 final project is separate from the two engineering projects in 36-614. This course grades the workforce-compensation report; the second mini-course covers databases, SQL, pipelines, and system handoffs.