Changelog

Source: NEWS.md

tidysoilinfiltration 1.1.0

New functions

• minidisk_conductivity() — convenience wrapper combining infiltration_vg_params(), optionally parameter_A_zhang(), and hydraulic_conductivity_minidisk() into a single pipeline step. Accepts method = "tabulated" (default, Decagon lookup) or method = "zhang" (analytical A via Zhang 1997). Reduces the Minidisk pipeline from 7 steps to 4 and eliminates the double group_by().
• ring_conductivity() — convenience wrapper combining infiltration_cumulative(), infiltration_rate(), and fit_infiltration_horton() into one call. Takes raw time–volume readings from a ponded ring infiltrometer and returns Horton parameters including .fc ≈ Kfs. Supports the same group_by() workflow as all other package functions.

Bug fixes

• infiltration_cumulative() and infiltration_rate() previously stripped group_by() grouping via an internal as_tibble() call, requiring users to re-group before calling fit_infiltration(). Both functions now preserve grouping through the pipeline, so a single group_by() is sufficient for the entire multi-site workflow.

---

tidysoilinfiltration 1.0.0

Initial release.

New functions

• infiltration_cumulative() — convert time–volume readings from Minidisk or ring infiltrometers to cumulative infiltration depth I(t) and √t. Fully group-aware.
• beerkan_cumulative() — convert BeerKan pour-event data (fixed volumes, infiltration times) to cumulative I(t). Group-aware via cumsum().
• infiltration_rate() — compute mean interval infiltration rates and midpoint times from cumulative data. Group-aware.
• fit_infiltration() — fit the Philip (1957) two-term polynomial I = C₂√t + C₁t by OLS. Returns C₁ (conductivity proxy) and C₂ (sorptivity proxy). Supports parallel fitting across groups.
• infiltration_vg_params() — look up Van Genuchten n, α, and A from the Decagon Devices (2005) table by texture class and suction. Pass suction = 0 for BeerKan (shape parameters only; .A is NA).
• parameter_A_zhang() — compute the Zhang (1997) A parameter analytically from n, α, suction, and disc radius. Fully vectorised.
• hydraulic_conductivity_minidisk() — compute K(h) = C₁ / A.
• fit_infiltration_horton() — fit the Horton (1940) exponential rate decay model by NLS. Returns fc ≈ Ksat, f₀, k.
• fit_infiltration_kostiakov() — fit the Kostiakov (1932) power model I = a t^b by NLS. Returns a, b.
• fit_best() — implement the BEST algorithm (Lassabatère et al., 2006). Estimates Ks, S, and α from BeerKan cumulative data using the Haverkamp et al. (1994) quasi-exact implicit model. Three fitting methods: "steady" (default), "slope", "intercept". Supports parallel fitting.

Datasets

• minidisk_vg_params — long-format tibble (96 rows × 5 columns) of Van Genuchten parameters n, α, and A for 12 USDA texture classes at 8 suction levels (0.5–7 cm). Source: Decagon Devices, Inc. (2005), via infiltrodiscR.

Notes

• All functions are pipe-compatible (data frame as first argument, returns tibble) and use tidy evaluation for column arguments.
• Fully vectorised backends — no R-level loops — suitable for raster-scale workflows at millions of cells.
• Parallel fitting via parallel::mclapply() on Unix-like systems (workers argument on all fit_* functions).