added pkgs to various notes

Author: ercbk

qmd/algorithms-ml.qmd

@@ -259,6 +259,7 @@
 ### Misc {#sec-alg-ml-knn-misc .unnumbered}
 
 -   Packages
+    -   [{]{style="color: #990000"}[bigKNN](https://cran.r-project.org/web/packages/bigKNN/index.html){style="color: #990000"}[}]{style="color: #990000"} - Exact Search and Graph Construction for 'bigmemory' Matrices
     -   [{]{style="color: #990000"}[kknn](https://cran.r-project.org/web/packages/kknn/index.html){style="color: #990000"}[}]{style="color: #990000"} - Weighted k-Nearest Neighbors for Classification, Regression and Clustering.
     -   [{]{style="color: #990000"}[LCCkNN](https://cran.r-project.org/web/packages/LCCkNN/index.html){style="color: #990000"}[}]{style="color: #990000"} - Adaptive k-Nearest Neighbor Classifier Based on Local Curvature Estimation
         -   Implements the kK-NN algorithm, an adaptive k-nearest neighbor classifier that adjusts the neighborhood size based on local data curvature
@@ -272,6 +273,7 @@
         -   It shows a preprocessing transformation that is performed before kNN to make it more efficient
         -   It might already be implemented in ANN algorithms
     -   Packages
+        -   [{]{style="color: #990000"}[bigANNOY](https://cran.r-project.org/web/packages/bigANNOY/index.html){style="color: #990000"}[}]{style="color: #990000"} - Approximate k-Nearest Neighbour Search for 'bigmemory' Matrices with Annoy
         -   [{]{style="color: goldenrod"}[faiss-gpu, cpu](https://faiss.ai/){style="color: goldenrod"}[}]{style="color: goldenrod"} - Efficient similarity search and clustering of dense vectors.
             -   Contains algorithms that search in sets of vectors of any size, up to ones that possibly do not fit in RAM.
             -   Contains supporting code for evaluation and parameter tuning.

qmd/econometrics-general.qmd

@@ -247,6 +247,7 @@
 
 -   Packages
 
+    -   [{]{style="color: #990000"}[catviz](https://cran.r-project.org/web/packages/catviz/index.html){style="color: #990000"}[}]{style="color: #990000"} - Visualizing Causal Assignment Trees for Callaway and Sant'Anna difference-in-differences (CSDiD) and doubly robust difference-in-difference-differences (DR-DDD) Designs
     -   [{]{style="color: #990000"}[did](https://bcallaway11.github.io/did/){style="color: #990000"}[}]{style="color: #990000"} - Computes ATE for DiD and allows for:
         -   More than two time periods
         -   Variation in treatment timing (i.e., units can become treated at different points in time)

qmd/llms-agent-chains.qmd

@@ -3,6 +3,9 @@
 ## Misc {#sec-llm-achains-misc .unnumbered}
 
 -   Packages
+    -   [{]{style="color: #990000"}[aisdk](https://cran.r-project.org/web/packages/aisdk/index.html){style="color: #990000"}[}]{style="color: #990000"} - A production-grade AI toolkit for R featuring a layered architecture (Specification, Utilities, Providers, Core), request interception support, robust error handling with exponential retry delays,
+        -   Support for multiple AI model providers ('OpenAI', 'Anthropic', etc.) and local small language model inference.
+        -   Includes a distributed 'MCP' ecosystem, multi-agent orchestration, progressive knowledge loading through skills, and a global skill store for sharing AI capabilities
     -   [{]{style="color: goldenrod"}[chatlas](https://posit-dev.github.io/chatlas/){style="color: goldenrod"}[}]{style="color: goldenrod"} - Posit's agent orchestration tool
     -   [{]{style="color: #990000"}[mini007](https://cran.r-project.org/web/packages/mini007/index.html){style="color: #990000"}[}]{style="color: #990000"} - Lightweight Framework for Orchestrating Multi-Agent Large Language Models
         -   Tools for creating agents with persistent state using R6 classes and the [{ellmer}]{style="color: #990000"}

qmd/mathematics-linear-algebra.qmd

@@ -3,6 +3,9 @@
 ## Misc {#sec-math-linalg-misc .unnumbered}
 
 -   Packages
+    -   [{]{style="color: #990000"}[gghinton](https://cran.r-project.org/web/packages/gghinton/index.html){style="color: #990000"}[}]{style="color: #990000"} - {ggplot2} extension for drawing Hinton diagrams,
+        -   A visualization technique for numerical matrices in which the area of each square is proportional to the magnitude of the corresponding entry
+        -   Useful for visualising PCA weight matrices, correlation matrices, and transition matrices.
     -   [{]{style="color: #990000"}[GPUmatrix](https://cran.r-project.org/web/packages/GPUmatrix/index.html){style="color: #990000"}[}]{style="color: #990000"} ([Github](https://github.com/ceslobfer/GPUmatrix)) - Mimics the behavior of [{Matrix}]{style="color: #990000"} and extends R to use the GPU for computations. It includes single(FP32) and double(FP64) precision data types, and provides support for sparse matrices. Able to use Tensorflow or Torch.
     -   [{]{style="color: #990000"}[Matrix](https://cran.r-project.org/web/packages/Matrix/index.html){style="color: #990000"}[}]{style="color: #990000"} - A rich hierarchy of sparse and dense matrix classes, including general, symmetric, triangular, and diagonal matrices with numeric, logical, or pattern entries.
         -   Efficient methods for operating on such matrices, often wrapping the 'BLAS', 'LAPACK', and 'SuiteSparse' libraries

qmd/missingness.qmd

@@ -10,6 +10,8 @@
     -   Model employs a uniform variable-selection mechanism across imputed datasets for prediction
 -   [{]{style="color: #990000"}[CoImp](https://cran.r-project.org/web/packages/CoImp/){style="color: #990000"}[}]{style="color: #990000"} - Parametric and Non-Parametric **Copula-Based** Imputation Methods
     -   The appropriate types of missingness are not explicitly mentioned in the docs, but based on the dgp functions, MCAR and MAR seem to the use cases.
+-   [{]{style="color: #990000"}[CompositionalNAimp](https://cran.r-project.org/web/packages/CompositionalNAimp/index.html){style="color: #990000"}[}]{style="color: #990000"} - Missing Value Imputation with **Compositional Data**
+    -   Uses the Jensen-Shannon divergence based k–NN and a–k–NN algorithms
 -   [{]{style="color: #990000"}[DebiasInfer](https://github.com/zhangyk8/Debias-Infer/tree/main/R_Package){style="color: #990000"}[}]{style="color: #990000"} ([Paper](https://arxiv.org/abs/2309.06429)) - Debiasing method for conducting valid inference on the **high-dimensional linear regression function with missing outcomes that are MAR**
     -   Combines a Lasso pilot estimate of the regression function with a bias correction term based on the weighted residuals of the Lasso regression. The weights depend on estimates of the missingness probabilities (propensity scores) and solve a convex optimization program that trades off bias and variance optimally.
     -   [{]{style="color: goldenrod"}[Debias-Infer](https://debias-infer.readthedocs.io/en/latest/index.html){style="color: goldenrod"}[}]{style="color: goldenrod"}

qmd/quarto-typst.qmd

@@ -9,6 +9,11 @@
 -   Packages
 
     -   [{]{style="color: #990000"}[r2typ](https://y-sunflower.github.io/r2typ/){style="color: #990000"}[}]{style="color: #990000"} - An R package for generating Typst markup
+    -   [{]{style="color: #990000"}[tyndling](https://cran.r-project.org/web/packages/tynding/index.html){style="color: #990000"}[}]{style="color: #990000"} - Provides bindings to the 'Typst' typesetting system, enabling users to compile 'Typst' documents directly from R.
+        -   The package interfaces with the 'Typst' 'Rust' library to render documents, making it possible to integrate 'Typst'-based workflows into R scripts, reports, and reproducible research pipelines.
+        -   Supports programmatic document generation, compilation, and output handling, facilitating seamless use of 'Typst' alongside tools such as knitr and Quarto
+        -   Similar to [{typr}]{style="color: #990000"} which compiles your document using the Typst/Quarto CLI, while [{tynding}]{style="color: #990000"} uses the Typst compiler itself via the Typst Rust library.
+        -   Portable in the sense that you don’t have to worry about installing Typst separately and/or adding it to your PATH.
     -   [{]{style="color: #990000"}[typstable](https://freierson.github.io/typstable/){style="color: #990000"}[}]{style="color: #990000"} - Produce publication-ready tables for Quarto documents targeting the Typst format
 
 -   Documentation

qmd/regression-survival.qmd

@@ -49,6 +49,7 @@ fig-cap-location: top
 -   [{]{style="color: #990000"}[jsurvival](https://www.serdarbalci.com/jsurvival/){style="color: #990000"}[}]{style="color: #990000"} - A comprehensive survival analysis module for jamovi that bridges the gap between **advanced statistical methods** and clinical research accessibility
     -   [Jamovi](https://www.jamovi.org/) is free, open source spreadsheet software
     -   **Extensive vignettes**
+-   [{]{style="color: #990000"}[mmbcv](https://cran.r-project.org/web/packages/mmbcv/index.html){style="color: #990000"}[}]{style="color: #990000"} - Computes **robust and bias-corrected sandwich variance** estimators for multi-state Cox models with clustered time-to-event data
 -   [{]{style="color: #990000"}[msm](https://github.com/chjackson/msm){style="color: #990000"}[}]{style="color: #990000"} - Multi-State Models
     -   Also see [{flexsurv}]{style="color: #990000"}
     -   Resources

qmd/shiny-general.qmd

@@ -60,6 +60,19 @@
 -   Comparison with other products ([article](https://www.appsilon.com/post/rshiny-vs-powerbi-vs-spotfire?utm_source=social&utm_medium=mastodon&utm_campaign=blog&utm_term=appsilon-account&utm_content=data-visualization))\
     ![](_resources/Shiny-General.resources/comp-shiny-powerbi-spotfire-1.png){.lightbox width="532"}
 -   Alternative Frameworks
+    -   [{]{style="color: #990000"}[RDesk](https://janakiraman-311.github.io/RDesk/){style="color: #990000"}[}]{style="color: #990000"} - A framework for building native Windows desktop applications with R. It turns your logic into a standalone .exe that runs on any machine—no R installation required by the end-user.
+        -   Use Cases
+            -   Data analysts building internal tools that cannot live on a server
+            -   Consultants distributing one-off analysis tools to clients
+            -   Teams replacing Excel macros with proper R-powered apps
+            -   Organisations that need offline, zero-IT-involvement deployment
+        -   Shiny better for: web applications, cross-platform needs, or real-time collaborative tools
+        -   Benefits
+            -   Zero-Port IPC: Native bidirectional pipes between R and the UI. No firewall issues or port conflicts.
+            -   Async by Default: Built-in background task processing via `mirai`. The UI never freezes, even during heavy R computations.
+            -   Portable Runtime: Packages a minimal R distribution into your `.exe`. Your users don’t need to install R.
+            -   Modern Web UI: Use HTML/JS/CSS for the interface while keeping 100% of your logic in R.
+            -   Professional Scaffolding: Generate dashboards with sidebar navigation, Dark Mode, and auto-wired charts in one command.
     -   [{]{style="color: #990000"}[htmxr](https://cran.r-project.org/web/packages/htmxr/index.html){style="color: #990000"}[}]{style="color: #990000"} ([website](https://htmx.org/)) - A lightweight framework for building server-driven web applications in 'R'. 'htmxr' combines the simplicity of 'htmx' for partial page updates with the power of 'plumber2' for non-blocking HTTP endpoints.
         -   Shiny maintains one R session per connected user. That session holds all the reactive state — inputs, outputs, intermediate computations — and Shiny decides when to re-evaluate which outputs. This model is powerful and ergonomic, but it means each user consumes a persistent R process.
 
@@ -68,7 +81,7 @@
         -   Summary
 
             |   | Shiny | htmxr |
-            |----|----|----|
+            |------------------------|------------------------|------------------------|
             | Communication | WebSocket (persistent connection) | HTTP (request/response) |
             | Paradigm | Reactive graph | Explicit HTTP requests |
             | UI updates | Shiny decides what to reload | You target the DOM precisely |

scrapsheet.qmd

@@ -335,6 +335,7 @@ $$
     -   [{]{style="color: #990000"}[fdars](https://cran.r-project.org/web/packages/fdars/index.html){style="color: #990000"}[}]{style="color: #990000"} - Written in Rust. Provides methods for functional data manipulation, depth computation, distance metrics, regression, and statistical testing. Supports both 1D functional data (curves) and 2D functional data (surfaces).
     -   [{]{style="color: #990000"}[fkcentroids](https://cran.r-project.org/web/packages/fkcentroids/index.html){style="color: #990000"}[}]{style="color: #990000"} - Functional K-Centroids Clustering Using Phase and Amplitude Components
     -   [{]{style="color: #990000"}[hdftsa](https://cran.r-project.org/web/packages/hdftsa/index.html){style="color: #990000"}[}]{style="color: #990000"} - High-Dimensional Functional Time Series Analysis
+    -   [{]{style="color: #990000"}[MFSD](https://cran.r-project.org/web/packages/MFSD/index.html){style="color: #990000"}[}]{style="color: #990000"} - Analysis of multivariate functional spatial data, including spectral multivariate functional principal component analysis and related statistical procedures
     -   [{]{style="color: #990000"}[mlr3fda](https://mlr3fda.mlr-org.com/){style="color: #990000"}[}]{style="color: #990000"} - fda in mlr3
     -   [{]{style="color: #990000"}[refund](https://cran.r-project.org/web/packages/refund/index.html){style="color: #990000"}[}]{style="color: #990000"} - Methods for regression for functional data, including function-on-scalar, scalar-on-function, and function-on-function regression.
     -   [{]{style="color: #990000"}[roahd](https://astamm.github.io/roahd/){style="color: #990000"}[}]{style="color: #990000"} - The Robust Analysis of High-dimensional Data package allows to use a set of statistical tools for the exploration and robustification of univariate and multivariate functional datasets through the use of depth-based statistical methods.