LoginSign Up
← Back to all sparks
F

fellingdater

ANALYTICS
Velocity0.0

A suite of functions designed to assist dendrochronologists in inferring estimates for felling dates, derived from dated tree-ring series.

github.com/ropensci/fellingdater ↗

Went from estimating felling dates to doing the crossdating that produces them.

dendrochronologycrossdatingarchaeologyropenscir-package
◆Current state
fellingdater estimates when a tree was felled from sapwood measurements, the core inference in dendrochronological dating of timber. Version 1.0.0 passed rOpenSci review with that scope, and the 2024 releases were mostly about the accompanying JOSS paper and user-supplied sapwood datasets. Version 1.2.0 changed the package's remit substantially, adding an entire trs_* family for tree-ring series handling: crossdating with multiple statistical measures, the Hollstein and Baillie-Pilcher t-value transformations, parallel variation percentages, synthetic series generation, and dated-series plotting.
◆Where it's heading
The package has expanded backwards along the workflow. It began at the last step — given dated series, estimate the felling date — and 1.2.0 added the step before it, establishing those dates by crossdating in the first place. Version 1.2.1 is early polish on that new surface: axis control, non-syntactic column names, encoding safety in read_fh(). The direction is a single package covering the chain from raw ring widths to a felling-date estimate.
◆Prediction
Expect the trs_* family to keep accumulating polish and additional crossdating statistics, since it is barely a year old and 1.2.1 was already fixing its plotting and top_n behaviour. Whether the two halves of the package get unified into one workflow interface is the open question the entries do not answer.

◆Recent moves

  1. 4mo ago

    Polish for the crossdating plots and file reader

    First round of fixes on the trs_* surface added in 1.2.0: configurable x-axis breaks and better y-axis scaling in trs_plot_dated(), safe handling of non-syntactic column names, and a correction to trs_crossdate() when top_n is active. read_fh() also normalises non-UTF-8 input rather than failing on it.

    View source ↗
  2. 1y ago

    Adds a full crossdating and tree-ring analysis toolkit

    ⚡ SPARK

    The release that widens the package from felling-date estimation to the crossdating that produces the dated series in the first place. It is the largest change in the package's history and defines what the 1.2.x line is now polishing.

    View source ↗
  3. 1y ago

    User-supplied sapwood data works across all functions

    Makes custom sapwood datasets usable consistently in fd_report(), sw_combine() and sw_sum(), backed by new input validation helpers for duplicate labels and dataset structure. Regional sapwood models vary enough that user-supplied data is the normal case, not the exception.

    View source ↗
  4. 1y ago

    Fixes fd_report() with user-defined sapwood files

    A single error fix for the user-supplied sapwood path, addressed more thoroughly two weeks later in 1.1.0. No other changes.

    View source ↗
  5. 2y ago

    JOSS paper accepted; citation updated

    Metadata only — the companion paper was accepted in the Journal of Open Source Software and the citation was updated. No code change, though for a research package the citable reference is part of what makes it adoptable.

    View source ↗
  6. 2y ago

    Adds a workflow vignette ahead of JOSS submission

    Documentation work in preparation for the JOSS paper, including a new article walking through the package workflow. No functional change.

    View source ↗