<?xml version="1.0" encoding="utf-8" ?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:r="https://r-universe.dev"><channel><title>itchyshin.r-universe.dev</title><link>https://itchyshin.r-universe.dev</link><description>Recent package updates in itchyshin</description><generator>R-universe</generator><image><url>https://github.com/itchyshin.png</url><title>R packages by itchyshin</title><link>https://itchyshin.r-universe.dev</link></image><lastBuildDate>Mon, 01 Jun 2026 21:54:40 GMT</lastBuildDate><item><title>[itchyshin] pigauto 0.10.0.9000</title><author>itchyshin@gmail.com (Shinichi Nakagawa)</author><description>Imputes missing species trait data for comparative
analyses by combining three sources of information:
phylogenetic similarity (closely related species share similar
traits), cross-trait correlations (observed traits inform
missing ones), and optional environmental covariates (climate,
habitat, geography). Handles continuous measurements, counts,
binary variables, ordered categories, unordered categories,
bounded proportions, zero-inflated counts, and compositional
multi-proportion data in a single call. The method blends a
phylogenetic baseline with a graph neural network correction; a
per-trait gate calibrated on held-out data ensures the network
only contributes when it improves on the baseline. Provides
conformal prediction intervals (95% coverage) for continuous,
count, and ordinal traits and supports Rubin's-rules multiple
imputation for downstream inference, including tree-uncertainty
propagation via posterior tree samples. Tested up to 10,000
species. Bundled datasets include a 300-species and a
9,993-species AVONET bird trait + BirdTree phylogeny subset.</description><link>https://github.com/r-universe/itchyshin/actions/runs/26787407268</link><pubDate>Mon, 01 Jun 2026 21:54:40 GMT</pubDate><r:package>pigauto</r:package><r:version>0.10.0.9000</r:version><r:status>success</r:status><r:repository>https://itchyshin.r-universe.dev</r:repository><r:upstream>https://github.com/itchyshin/pigauto</r:upstream><r:article><r:source>common-pitfalls.Rmd</r:source><r:filename>common-pitfalls.html</r:filename><r:title>Common pitfalls / FAQ</r:title><r:created>2026-05-10 11:06:34</r:created><r:modified>2026-05-12 17:26:16</r:modified></r:article><r:article><r:source>getting-started.Rmd</r:source><r:filename>getting-started.html</r:filename><r:title>Getting started with pigauto: Phylogenetic Imputation via Graph AUTO-encoders</r:title><r:created>2026-04-06 22:25:32</r:created><r:modified>2026-05-12 17:26:16</r:modified></r:article><r:article><r:source>gnn-architecture.Rmd</r:source><r:filename>gnn-architecture.html</r:filename><r:title>GNN architecture and the math behind pigauto</r:title><r:created>2026-05-18 15:46:53</r:created><r:modified>2026-05-26 22:33:09</r:modified></r:article><r:article><r:source>mixed-types.Rmd</r:source><r:filename>mixed-types.html</r:filename><r:title>Mixed-Type Trait Imputation</r:title><r:created>2026-04-07 11:50:29</r:created><r:modified>2026-05-12 17:26:16</r:modified></r:article><r:article><r:source>tree-uncertainty.Rmd</r:source><r:filename>tree-uncertainty.html</r:filename><r:title>Propagating Tree Uncertainty</r:title><r:created>2026-04-18 14:01:48</r:created><r:modified>2026-05-12 17:26:16</r:modified></r:article></item></channel></rss>