Research Contributions Map
How The Corpus Fits Together
This site collects a research and software corpus rather than a single book companion page. The most useful way to navigate it is by topic.
| Topic | Primary Sources | Software Entry Points |
|---|---|---|
| Mixed-data kernel smoothing | Princeton theory, Cambridge advanced treatment | np, npRmpi |
| Kernel density, distribution, and quantiles | Princeton, Cambridge | npudens, npudist, npcdens, npcdist, npqreg |
| Conditional mean and regression | Princeton, Cambridge | npreg, npregbw, predict, plot |
| Semiparametric models | Cambridge, Princeton | npplreg, npindex, npscoef |
| Endogeneity and IV-style nonparametric regression | Cambridge, Princeton | npregiv routes where applicable |
| Regression splines and categorical predictors | Oxford spline primer, crs article |
crs, crs-package documentation |
| Reproducible computation in R | Oxford, Cambridge setup chapters | R, RStudio, TeX, Git, Quarto/bookdown workflows |
| Robust inference, resampling, model uncertainty | Oxford | boot, model averaging materials, legacy ma |
| Large-data and parallel execution | Package vignettes, npRmpi examples |
npRmpi, MPI profile/attach/session workflows |
Reading Paths
For audience-specific routes through the corpus, see Reader Pathways.
For Nonparametric Econometrics
Start with Li and Racine (2007) for the broad theoretical foundation, then use Racine (2019a) for a replicable R-centered treatment of advanced topics. The np package article (2008) and package vignettes connect these methods to day-to-day R workflows.
For Reproducible Econometrics
Start with Racine (2019b) for reproducible R workflows, robust inference, time series, model uncertainty, and related computational tools. Use the Downloads and Teaching page for public errata and companion code.
For Package Users
Use the R Packages page for current package links, package versions, CRAN/manual/vignette pointers, and GitHub source repositories. Use the Gallery of Code for worked examples and current package-oriented snippets.
Public Versus Instructor Materials
The public site should surface durable public materials: publisher links, errata, companion code, public manuals, vignettes, package articles, and selected solution material that is already intended for public use.
Instructor-only material, including full solutions manuals, exams, assignments, and private slide decks for adopting instructors, should be described here but not published directly unless explicitly intended for public distribution.
Site Maintenance Guidance
For future updates, this Research site should stay organized around durable reader questions:
- Which book or package should I start with?
- Where are the public companion files?
- Which software implements a given method family?
- Which materials are public, and which require an instructor request?
- How does this site relate to the package Gallery?
The Gallery can stay code-first; this site should stay corpus-first.