Install R and RStudio, configure your environment, and verify with a test script.

R objects, vectors, and data frames from scratch. Arithmetic operators, summary statistics, base-R graphics, and simple linear models — demonstrated with a lake temperature profile case study.

Organize scripts, data, and outputs using RStudio Projects and a consistent folder hierarchy.

Weave prose, code, and output into reproducible reports. YAML headers, chunk options, rendering.

Reading CSV, Excel, and field-data files. Handling encoding issues and common pitfalls.

Measures of center and spread, frequency tables, base-R and dplyr summaries.

Core ggplot2 grammar: geoms, aesthetics, facets, themes, and export-ready figures.

Every statistical test you will encounter in environmental science fits into one of four frameworks determined by a single question: what type of variable is your outcome (dependent), and what type of variable is your predictor (independent)?

Every classical statistical test — t-test, ANOVA, regression — is a formalization of the same underlying question: is the pattern I see in my data larger than what I would expect from chance variation alone?

t-tests, Wilcoxon tests, proportions tests. Effect sizes, p-values, common misconceptions.

Fitting, diagnosing, and interpreting OLS models with lm(). Residual plots and leverage.

One-way and two-way ANOVA, post-hoc tests, and visualizing group differences.

Chi-square, Fisher’s exact, and correlation tests. Choosing the right test for categorical and continuous variables, effect sizes (Cramér’s V, odds ratio, r), and assumption checking.

Modeling binary outcomes — presence/absence, exceedance, survival — with logistic regression. Odds ratios, ROC curves, and model diagnostics.

Pearson and Spearman correlation, scatterplot matrices, and interpreting correlation structure.

Multiple predictors, interactions, model selection (AIC), and GLMs for count and binary data.

Random intercepts and slopes with lme4. Nested and crossed random effects for field designs.

Penalized regression for automatic variable selection. Fit, cross-validate, and interpret LASSO models with glmnet on environmental water quality data. Includes ridge and elastic net comparison.

Learn to handle environmental data that instruments can’t fully measure — and why faking it breaks your statistics.

Decomposition, autocorrelation, ARIMA, and sensor-data workflows with lubridate and tsibble.

Time-to-event analysis with Kaplan–Meier curves, log-rank tests, and Cox proportional hazards models. Applied to algae growth experiments and right-censored environmental data.

Resampling-based inference, confidence intervals, and randomization tests.

Importing shapefiles, CRS transformations, spatial joins, and choropleth maps with sf.

PCA, NMDS, and RDA for community and environmental gradient data using vegan.

for loops, the base-R apply family (lapply, sapply, vapply, mapply), and purrr::map() compared side-by-side. Replace repetitive loops with cleaner, type-safe pipelines.

dplyr verbs, tidyr pivoting, and pipe-based workflows for cleaning field and lab data.

Publication-quality tables with kableExtra, gt, and flextable. Inline R in RMarkdown.

Version-control basics, committing from RStudio, branching, and submitting via pull request.

Lock package versions and share reproducible environments across machines.

Decode common R error messages, trace the source of warnings, and build a systematic debugging strategy.

Parse, manipulate, and plot date-time data with lubridate. Covers POSIXct vs Date, time zones, floor_date(), and sensor timestamp correction.

Overview of all EA030 projects this semester — Regional Climate Trends, pHake Lake, Compost Maturity, Inland Empire Air Quality, and Surface Water Quality.

Thirty years of daily fine particulate matter (PM2.5) for a US county. Download EPA AQS data, clean monitoring records, visualize long-term trends, compute monthly means with 95% CIs by decade, and evaluate exceedances of the 24-hr NAAQS (35 µg/m³). Includes environmental justice discussion.

Deploy Raspberry Pi + Enviro+ sensors to monitor PM1/PM2.5/PM10 across campus. Covers clock-correction for RTC-less Pis, QC, diurnal patterns, Wilcoxon/Kruskal-Wallis tests, OLS regression, and mixed-effects models (lmer + nlme AR(1)) to handle pseudoreplication and temporal autocorrelation.

Use Solvita paddle tests to measure compost stability via CO₂ and NH₃ emissions. Import class data, visualise maturity scores across treatments, and apply t-tests, one-way ANOVA, and Tukey HSD to compare compost recipes. Includes assumption checking and a non-parametric Kruskal-Wallis alternative.

Importing and cleaning soil survey data, fitting depth profiles, and comparing land-use treatments.

.

Time-series of field sensor data, detecting exceedances, and visualising spatial variation.

.

.

Species richness, rarefaction curves, NMDS ordination, and PERMANOVA for plant and bird communities in the Pomona College Wash. Uses vegan.

Camera traps record who was present, when, and what they were doing without a researcher present. This tutorial walks the complete analytical pipeline — from raw Excel export through cleaning, independent-event definition, descriptive ethograms, and four quantitative methods: chi-square/Fisher tests of diel behavior shifts, kernel-density activity overlap, rarefaction-corrected diversity, and single-season occupancy modeling.

Measuring tone in environmental news with dictionary-based, aspect-based, and transformer methods. Build validated time-series sentiment indices from a text corpus using quanteda, sentimentr, tidytext, and the text package.

A 16-script ensemble forecasting pipeline for Brent crude and California gasoline. ARIMA-X, Prophet, LASSO, bias-penalized weighting, spot anchoring, GAM spread model, and diagnostic-driven iteration across multiple versions.

Comparing sun and shade temperatures in parks and public spaces across Southern California using paired iButton sensors. Analyze shade treatment effects with a linear mixed-effects model.

Raster data from Landsat, land-cover classification, and surface temperature analysis.

Hadley Wickham & Mine Cetinkaya-Rundel. The canonical tidyverse reference, freely online.

Dense, searchable reference for base-R syntax — useful as a cheat-sheet during lab sessions.

Full function reference for every geom, stat, scale, and theme component in ggplot2.

Official one-page PDF cheatsheets for ggplot2, dplyr, tidyr, RMarkdown, and more.