sddr 0.1.1
CRAN release: 2026-07-19
- Initial CRAN release.
- Removed an unresolvable DOI from the Quah (1993) reference.
-
markov(): classic discrete-time, first-order Markov transition estimation from tidy long-formatid/time/valuepanel data, with quantile-based (fixed or per-period) class discretisation. -
full_rank_markov()andgeo_rank_markov(): rank-based Markov chains (ranks-as-states, and units-exchanging-rank-positions) — no binning required. -
lisa_markov(): Markov chain over the four Moran-scatterplot quadrants (HH/LH/LL/HL), capturing the joint dynamics of a unit and its neighbourhood. -
spatial_markov(): spatial Markov chain (Rey 2001) with transition matrices conditioned on the spatial-lag class of the neighbourhood. Accepts either a spatial weights matrix (lag computed internally) or a precomputed lag column, and explicitbreaks/lag_breakscut points. With matching cut points it reproduces PySALgiddy’s spatial Markov matrices to machine precision. -
steady_state(): ergodic / stationary distribution of a transition matrix or ansddr_markovobject. -
mfpt(): mean first passage times (Kemeny-Snell fundamental matrix), with mean recurrence times on the diagonal. -
sojourn_time(): expected persistence time in each class. -
tau(): Kendall’s tau rank correlation (positional / exchange mobility), with concordant/discordant counts and an asymptotic p-value. -
theta(): Theta rank-mobility statistic (Rey 2004) decomposing rank change by regime. -
tau_local(): per-observation (local) Kendall’s tau decomposition. -
mobility(): Markov mobility indices (Prais, determinant, second-eigenvalue L2, and the Shorrocks B1/B2 indices). -
print()methods forsddr_markovandsddr_spatial_markov. -
spatial_markov()andlisa_markov()now accept spatial weights as anlistwornbobject (e.g. built from ansflayer), in addition to a plain weights matrix. ‘spdep’ is an optional (Suggests) dependency. - Hex logo, pkgdown site configuration, and GitHub Actions R-CMD-check CI.