Articles
Guides
- Your first capsule: a published table, analysed and recorded
- Building reproducible data capsules
- Provenance you can verify
- Look at the data first
- Categorical integrity: labels that cannot be swapped quietly
- Has the data changed?
- Year-over-year change, and the three ways it goes wrong
- Statistics for a published administrative table
- Plotting results
Cryptography and assurance
- Post-quantum signatures: ML-DSA, SLH-DSA and XMSS
- Key encapsulation: ML-KEM and HQC
- Hashing, keyed digests, key derivation and Merkle trees
- Attestations, signed bundles and tamper-evident chains
- Side-channel assurance: constant time, timing tests, masking and simulated power analysis
- JSON without jsonlite: parsing, encoding and exact serialisation
- Language models (your own key) and the signed services document
Statistics for published tables
- Trends and change in short count series
- Small-area rates: expected counts, shrinkage, funnels and spatial autocorrelation
- Complex survey estimation
- Hawkes processes: self-exciting event series
- Banded tables and concentration: from '6 to 10' to a Gini
- Stock and flow: population, admissions and length of stay
- Falsification controls and power curves: is the statistic real?
- Air pollution: dispersion, health burden, equity, and the footprint of the computation
- Causal effects with cross-fitted double machine learning
- Courts and crime: flow, time, outcomes, place and case mix