From 070931c6f3a58764d2c8c13071979cfdc0bd0a2d Mon Sep 17 00:00:00 2001 From: Seongho Bae Date: Wed, 2 Sep 2026 07:24:16 +0900 Subject: [PATCH 1/4] docs(pages): refresh kaefa public landing page --- index.md | 57 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 57 insertions(+) create mode 100644 index.md diff --git a/index.md b/index.md new file mode 100644 index 0000000..fe41936 --- /dev/null +++ b/index.md @@ -0,0 +1,57 @@ +# kaefa + +[![Ask DeepWiki](https://deepwiki.com/badge.svg)](https://deepwiki.com/ContextualWisdomLab/kaefa) + +**kaefa** is an R package for automated exploratory factor analysis (AEFA). It helps researchers explore uncertain factor structures, compare candidate models, evaluate item fit, and iterate toward better-fitting solutions, including workflows for complex and multilevel data. + +## Product surfaces + +- **kaefa-core** — the statistical engine behind `aefa()` and `engineAEFA()`, including candidate-model search, information-criterion selection, item-fit evaluation, and theta-prior utilities. +- **kaefa-studio** — the bundled Shiny interface launched with `launchAEFA()` for researchers who prefer an interactive workflow. +- **Remote execution** — optional worker initialization through `aefaInit()` for analyses that outgrow a local workstation. + +The package remains R/`mirt` based. Current architecture and supported boundaries are documented in the repository rather than inferred from experimental branches. + +## How AEFA works + +The current AEFA workflow explores multiple candidate factor structures and item-response models, selects among fitted candidates using information criteria, checks item-level fit, removes poorly fitting items one at a time when appropriate, and re-estimates until the search converges. AIC is the default model-selection criterion; AICc, BIC, sample-size-adjusted BIC, and posterior DIC when a fitted model actually provides it are also supported. + +## Install + +Install the organization-owned repository directly from GitHub: + +```r +# install.packages("devtools") +devtools::install_github("ContextualWisdomLab/kaefa") +``` + +Then run a basic analysis: + +```r +library(kaefa) +fit <- kaefa::aefa(mirt::Science) +fit +``` + +For the interactive interface: + +```r +library(kaefa) +launchAEFA() +``` + +## Documentation and onboarding + +- [Repository README](https://github.com/ContextualWisdomLab/kaefa/blob/develop/README.md) — installation, examples, remote execution, workload sizing, Shiny usage, and quality information. +- [Architecture](https://github.com/ContextualWisdomLab/kaefa/blob/develop/ARCHITECTURE.md) — runtime flow, product boundaries, repository layout, and quality gates. +- [Source repository](https://github.com/ContextualWisdomLab/kaefa) — issues, pull requests, releases, code, and current development activity. +- [Releases](https://github.com/ContextualWisdomLab/kaefa/releases) — packaged release history when available. +- [Ask DeepWiki](https://deepwiki.com/ContextualWisdomLab/kaefa) — repository-aware questions about the codebase and documentation. + +## Research and model-evidence boundary + +kaefa is a research-oriented statistical package. Model fit, convergence, item-fit diagnostics, and parameter-recovery evidence should be interpreted in the context of the data-generating process and the chosen IRT/factor model. New claims about supported model classes or recovery quality belong in reviewed code, tests, and traceability documentation before they are presented here as released capability. + +## License + +kaefa is distributed under the GNU General Public License v3.0. See the repository for the authoritative license text and current source. From 0fb2842b356d5c11929a3fe70797a2b5a4b38985 Mon Sep 17 00:00:00 2001 From: Seongho Bae Date: Wed, 2 Sep 2026 07:24:21 +0900 Subject: [PATCH 2/4] docs(pages): describe current kaefa product --- _config.yml | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/_config.yml b/_config.yml index c419263..c1162ad 100644 --- a/_config.yml +++ b/_config.yml @@ -1 +1,3 @@ -theme: jekyll-theme-cayman \ No newline at end of file +title: kaefa +description: Automated exploratory factor analysis for R, with model search, item-fit diagnostics, and multilevel workflows. +theme: jekyll-theme-cayman From c82ff0d6e2dfb02047126489edb9d880ba87fe4a Mon Sep 17 00:00:00 2001 From: Seongho Bae Date: Wed, 2 Sep 2026 09:15:05 +0900 Subject: [PATCH 3/4] docs(pages): wrap landing page for markdownlint --- index.md | 52 +++++++++++++++++++++++++++++++++++++++------------- 1 file changed, 39 insertions(+), 13 deletions(-) diff --git a/index.md b/index.md index fe41936..f067ad9 100644 --- a/index.md +++ b/index.md @@ -2,19 +2,33 @@ [![Ask DeepWiki](https://deepwiki.com/badge.svg)](https://deepwiki.com/ContextualWisdomLab/kaefa) -**kaefa** is an R package for automated exploratory factor analysis (AEFA). It helps researchers explore uncertain factor structures, compare candidate models, evaluate item fit, and iterate toward better-fitting solutions, including workflows for complex and multilevel data. +**kaefa** is an R package for automated exploratory factor analysis (AEFA). +It helps researchers explore uncertain factor structures, compare candidate +models, evaluate item fit, and iterate toward better-fitting solutions, +including workflows for complex and multilevel data. ## Product surfaces -- **kaefa-core** — the statistical engine behind `aefa()` and `engineAEFA()`, including candidate-model search, information-criterion selection, item-fit evaluation, and theta-prior utilities. -- **kaefa-studio** — the bundled Shiny interface launched with `launchAEFA()` for researchers who prefer an interactive workflow. -- **Remote execution** — optional worker initialization through `aefaInit()` for analyses that outgrow a local workstation. +- **kaefa-core** — the statistical engine behind `aefa()` and `engineAEFA()`, + including candidate-model search, information-criterion selection, item-fit + evaluation, and theta-prior utilities. +- **kaefa-studio** — the bundled Shiny interface launched with `launchAEFA()` + for researchers who prefer an interactive workflow. +- **Remote execution** — optional worker initialization through `aefaInit()` + for analyses that outgrow a local workstation. -The package remains R/`mirt` based. Current architecture and supported boundaries are documented in the repository rather than inferred from experimental branches. +The package remains R/`mirt` based. Current architecture and supported +boundaries are documented in the repository rather than inferred from +experimental branches. ## How AEFA works -The current AEFA workflow explores multiple candidate factor structures and item-response models, selects among fitted candidates using information criteria, checks item-level fit, removes poorly fitting items one at a time when appropriate, and re-estimates until the search converges. AIC is the default model-selection criterion; AICc, BIC, sample-size-adjusted BIC, and posterior DIC when a fitted model actually provides it are also supported. +The current AEFA workflow explores multiple candidate factor structures and +item-response models, selects among fitted candidates using information +criteria, checks item-level fit, removes poorly fitting items one at a time +when appropriate, and re-estimates until the search converges. AIC is the +default model-selection criterion; AICc, BIC, sample-size-adjusted BIC, and +posterior DIC when a fitted model actually provides it are also supported. ## Install @@ -42,16 +56,28 @@ launchAEFA() ## Documentation and onboarding -- [Repository README](https://github.com/ContextualWisdomLab/kaefa/blob/develop/README.md) — installation, examples, remote execution, workload sizing, Shiny usage, and quality information. -- [Architecture](https://github.com/ContextualWisdomLab/kaefa/blob/develop/ARCHITECTURE.md) — runtime flow, product boundaries, repository layout, and quality gates. -- [Source repository](https://github.com/ContextualWisdomLab/kaefa) — issues, pull requests, releases, code, and current development activity. -- [Releases](https://github.com/ContextualWisdomLab/kaefa/releases) — packaged release history when available. -- [Ask DeepWiki](https://deepwiki.com/ContextualWisdomLab/kaefa) — repository-aware questions about the codebase and documentation. +- [Repository README](https://github.com/ContextualWisdomLab/kaefa/blob/develop/README.md) + — installation, examples, remote execution, workload sizing, Shiny usage, + and quality information. +- [Architecture](https://github.com/ContextualWisdomLab/kaefa/blob/develop/ARCHITECTURE.md) + — runtime flow, product boundaries, repository layout, and quality gates. +- [Source repository](https://github.com/ContextualWisdomLab/kaefa) — issues, + pull requests, releases, code, and current development activity. +- [Releases](https://github.com/ContextualWisdomLab/kaefa/releases) — packaged + release history when available. +- [Ask DeepWiki](https://deepwiki.com/ContextualWisdomLab/kaefa) — + repository-aware questions about the codebase and documentation. ## Research and model-evidence boundary -kaefa is a research-oriented statistical package. Model fit, convergence, item-fit diagnostics, and parameter-recovery evidence should be interpreted in the context of the data-generating process and the chosen IRT/factor model. New claims about supported model classes or recovery quality belong in reviewed code, tests, and traceability documentation before they are presented here as released capability. +kaefa is a research-oriented statistical package. Model fit, convergence, +item-fit diagnostics, and parameter-recovery evidence should be interpreted in +the context of the data-generating process and the chosen IRT/factor model. +New claims about supported model classes or recovery quality belong in +reviewed code, tests, and traceability documentation before they are presented +here as released capability. ## License -kaefa is distributed under the GNU General Public License v3.0. See the repository for the authoritative license text and current source. +kaefa is distributed under the GNU General Public License v3.0. See the +repository for the authoritative license text and current source. From 86b94312498f3bffc6dbd63bd9d538c9183d5358 Mon Sep 17 00:00:00 2001 From: Seongho Bae Date: Wed, 2 Sep 2026 09:15:54 +0900 Subject: [PATCH 4/4] docs(pages): satisfy line-length lint --- index.md | 16 +++++++++++----- 1 file changed, 11 insertions(+), 5 deletions(-) diff --git a/index.md b/index.md index f067ad9..3025fa8 100644 --- a/index.md +++ b/index.md @@ -1,6 +1,11 @@ # kaefa -[![Ask DeepWiki](https://deepwiki.com/badge.svg)](https://deepwiki.com/ContextualWisdomLab/kaefa) + + Ask DeepWiki + **kaefa** is an R package for automated exploratory factor analysis (AEFA). It helps researchers explore uncertain factor structures, compare candidate @@ -56,10 +61,11 @@ launchAEFA() ## Documentation and onboarding -- [Repository README](https://github.com/ContextualWisdomLab/kaefa/blob/develop/README.md) - — installation, examples, remote execution, workload sizing, Shiny usage, - and quality information. -- [Architecture](https://github.com/ContextualWisdomLab/kaefa/blob/develop/ARCHITECTURE.md) +- [Repository README](https://github.com/ContextualWisdomLab/kaefa) — + installation, examples, remote execution, workload sizing, Shiny usage, and + quality information. +- **Architecture** + — runtime flow, product boundaries, repository layout, and quality gates. - [Source repository](https://github.com/ContextualWisdomLab/kaefa) — issues, pull requests, releases, code, and current development activity.