LoginSign Up
← Back to all sparks
R

reda

ANALYTICS
Velocity0.0

R package reda by wenjie2wang — release notes from GitHub.

github.com/wenjie2wang/reda ↗

A mature recurrent-event toolkit in careful maintenance, shedding weight rather than adding surface.

survival-analysisrecurrent-eventsmaintenance-modecranr-package
◆Current state
reda provides nonparametric mean cumulative function estimation, gamma-frailty rate regression, and event-data simulation for recurrent-event survival analysis. The core API settled at 0.5.0 when Recur() replaced Survr() and the MCF internals moved to C++. Everything since has been consolidation: small argument additions, method completions, and CRAN hygiene.
◆Where it's heading
The last three releases contain no new modelling capability at all — a dependency reshuffle, a test-example correction, and a print-order fix. The package is being kept installable and correct rather than extended. Its tightest coupling is to splines2, a sibling package from the same maintainer, which supplies the derivative machinery reda depends on.
◆Prediction
Expect continued small-cadence CRAN-compliance releases tracking ggplot2 and splines2 changes. The entries show no in-progress feature work, so a substantive release would have to arrive without warning from this feed.

◆Recent moves

  1. 11mo ago

    ggplot2 and grDevices demoted to Suggests

    ggplot2 and grDevices drop from hard dependencies to Suggests, since only plot() methods need them. Standard CRAN weight-trimming that leaves the modelling surface untouched.

    View source ↗
  2. 1y ago

    Test example and documentation typo fixes

    A test example correction and a documentation typo. Nothing here reaches the user's model code.

    View source ↗
  3. 4y ago

    simEvent() gains a user-supplied rate bound

    simEvent() gains rhoMax, letting users supply an upper bound on the rate function when the internal heuristic cannot find one. A small but real escape hatch for simulation cases that previously just failed.

    View source ↗
  4. 5y ago

    summary() method completes the Recur object

    A summary() method and a time_class slot complete the Recur() object introduced at 0.5.0, alongside fixes to empty MCF estimates and index slots. This is the release that made the post-Survr API usable in practice.

    View source ↗
  5. 5y ago

    Derivatives sourced from splines2's deriv method

    Derivatives now come from splines2's deriv method instead of attribute extraction. An internal swap onto the maintainer's sibling package, invisible at the call site.

    View source ↗
  6. 6y ago

    MCF estimates can now skip variance computation

    A variance = "none" option lets users compute sample MCF estimates without the variance calculation. Useful where only the point estimate matters and the variance step dominates runtime.

    View source ↗