The numbered package vignettes are the canonical NNS example curriculum. They define the statistical narrative, section order, datasets, and intended interpretation for companion implementations. The Python package follows this sequence and treats R as the source of truth.
| # | Topic | R vignette | Python companion |
|---|---|---|---|
| 01 | Overview | Overview | 01_overview.py |
| 02 | Partial Moments | Partial Moments | 02_partial_moments.py |
| 03 | Correlation and Dependence | Correlation and Dependence | 03_correlation_and_dependence.py |
| 04 | Normalization and Rescaling | Normalization and Rescaling | 04_normalization_and_rescaling.py |
| 05 | Sampling and Simulation | Sampling and Simulation | 05_sampling_and_simulation.py |
| 06 | Comparing Distributions | Comparing Distributions | 06_comparing_distributions.py |
| 07 | Clustering and Regression | Clustering and Regression | 07_clustering_and_regression.py |
| 08 | Classification | Classification | 08_classification.py |
| 09 | Forecasting | Forecasting | 09_forecasting.py |
Canonical changes should be made in the R vignette first. A companion port may use language-appropriate syntax and containers, but should preserve the same statistical demonstration and interpretation.
Read example corrections and reproducibility before using the historical applied studies. The PDF originals are retained; their current executable companion and source-provenance notes are linked there. The earlier interval-benchmark ranking and unsupported comparative accuracy claims have been withdrawn. Current results must be regenerated under the stated validation protocol.
The following material applies NNS to specific problems, comparisons, and research questions. These examples are useful applications, but they do not replace the numbered package curriculum.
The applied examples are demonstrations rather than exhaustive proofs. See the book and the research papers for the underlying arguments.