{"Package":"caugi","Title":"Causal Graph Interface","Version":"1.0.0","Authors@R":"c(\nperson(\ngiven = \"Frederik\",\nfamily = \"Fabricius-Bjerre\",\nemail = \"frederik@fabriciusbjerre.dk\",\nrole = c(\"aut\", \"cre\", \"cph\")\n),\nperson(\ngiven = \"Johan\",\nfamily = \"Larsson\",\nrole = \"aut\",\nemail = \"johan@jolars.co\",\ncomment = c(ORCID = \"0000-0002-4029-5945\")\n),\nperson(\ngiven = \"Michael\",\nfamily = \"Sachs\",\nrole = \"aut\",\nemail = \"sachsmc@gmail.com\",\ncomment = c(ORCID = \"0000-0002-1279-8676\")\n)\n)","Description":"Create, query, and modify causal graphs. 'caugi' (Causal Graph\nInterface) is a causality-first, high performance graph package that\nprovides a simple interface to build, structure, and examine causal\nrelationships.","License":"MIT + file LICENSE","Language":"en-US","URL":"https://caugi.org/","BugReports":"https://github.com/frederikfabriciusbjerre/caugi/issues","Depends":"R (>= 4.2)","Imports":"data.table, fastmap, grid, S7, stats, methods","Suggests":"bnlearn, dagitty, devtools, ggm, graph, gRbase, igraph,\njsonlite, knitr, MASS, Matrix, rextendr, rmarkdown, testthat","VignetteBuilder":"knitr","Config/rextendr/version":"0.4.2","Encoding":"UTF-8","RoxygenNote":"7.3.3","SystemRequirements":"Cargo (Rust's package manager), rustc >= 1.80.0, xz","Config/Needs/website":"rmarkdown","NeedsCompilation":"yes","Packaged":"2026-01-22 10:07:22 UTC; jola","Author":"Frederik Fabricius-Bjerre [aut, cre, cph],\nJohan Larsson [aut] (ORCID: ),\nMichael Sachs [aut] (ORCID: )","Maintainer":"Frederik Fabricius-Bjerre ","Repository":"CRAN","Date/Publication":"2026-01-22 11:00:13 UTC","repoType":"cran","tarballUrl":"ftp://cran.r-project.org/pub/R/src/contrib/caugi_1.0.0.tar.gz","jsonAuthors":[{"name":"Frederik Fabricius-Bjerre","email":"frederik@fabriciusbjerre.dk","maintainer":true},{"name":"Johan Larsson","email":"johan@jolars.co","maintainer":false},{"name":"Michael Sachs","email":"sachsmc@gmail.com","maintainer":false}],"readme":"\n\n\n# caugi

\n\n\n\n[![R-CMD-check](https://github.com/frederikfabriciusbjerre/caugi/actions/workflows/R-CMD-check.yaml/badge.svg)](https://github.com/frederikfabriciusbjerre/caugi/actions/workflows/R-CMD-check.yaml)\n[![Code\ncoverage](https://codecov.io/gh/frederikfabriciusbjerre/caugi/graph/badge.svg)](https://app.codecov.io/gh/frederikfabriciusbjerre/caugi)\n[![CRAN\ndownloads](https://cranlogs.r-pkg.org/badges/caugi)](https://cran.r-project.org/package=caugi)\n[![extendr](https://img.shields.io/badge/extendr-%5E0.8.1-276DC2)](https://extendr.rs/extendr/extendr_api/)\n\n\n> **Causal Graph Interface (for R)** — a fast and flexible toolbox for\n> building, coercing and analyzing causal graphs.\n\n## What is `caugi`?\n\n`caugi` (pronounced “corgi”) stands for **Causal Graph Interface**. It\nis a *causality-first* graph package that focuses on performance and\nflexibility. If you are developing scripts or algorithms in the field of\ncausality or if you are learning about causal graphs for the first time,\n`caugi` is made for you.\n\n## Installation\n\nYou can install the development version of `caugi` from GitHub with:\n\n``` r\npak::pak(\"frederikfabriciusbjerre/caugi\")\n```\n\nor directly from CRAN with:\n\n``` r\ninstall.packages(\"caugi\")\n```\n\n### Installing Rust\n\nRust is required for `caugi` to work. If you don’t have Rust installed,\nvisit [rustup.rs](https://rustup.rs/) for installation instructions\nappropriate for your platform.\n\nAlternatively, you may install Rust from your OS package manager.\n\n## The basic object: `caugi`\n\nA `caugi` graph object is the bread and butter of the `caugi` package.\nIt is easy to create, query, and modify.\n\n``` r\nlibrary(caugi)\n```\n\nYou can create simple graphs as well as a number of predefined graph\nclasses. Currently, we support `\"UNKNOWN\"`, `\"DAG\"`, `\"PDAG\"`, `\"ADMG\"`,\nand `\"UG\"`. We plan on supporting several other causal graph types in\nfuture releases, such as `\"PAG\"`, `\"MAG\"`, and `\"SWIG\"`.\n\n``` r\n# a tiny DAG\ncg <- caugi(\n A %-->% B + C,\n B %-->% D,\n C %-->% D,\n class = \"DAG\"\n)\n\nplot(cg)\n```\n\n![](man/figures/README-unnamed-chunk-5-1.png)\n\n### Edge operators\n\nThe available edges in `caugi` are listed below:\n\n- `%-->%` (directed)\n- `%---%` (undirected)\n- `%<->%` (bidirected)\n- `%o->%` (partially directed)\n- `%o--%` (partially undirected)\n- `%o-o%` (partial)\n\nYou can register more types with `register_caugi_edge()`, if you find\nthat you need a more expressive set of edges. For example, if you want\nto represent a directed edge in the reverse direction, you can do so\nlike this:\n\n``` r\nregister_caugi_edge(\n glyph = \"<--\",\n tail_mark = \"arrow\",\n head_mark = \"tail\",\n class = \"directed\",\n symmetric = FALSE\n)\n\ncaugi(A %-->% B, B %<--% C, class = \"DAG\")\n#> \n#> graph_class: DAG\n#> nodes: A, B, C\n#> edges: A-->B, B<--C\n\n# reset the registry to default with original edges\nreset_caugi_registry()\n```\n\nWe expect this feature to be needing further polishing in future\nreleases, and we would love your input if you use this feature!\n\n## Querying and metrics\n\n`caugi` provides a number of functions to query and analyze `caugi`\nobjects. Some of the available functions are:\n\n- Relational queries, such as `parents()`, `ancestors()`, `neighbors()`,\n and more.\n- Structural queries, such as `is_acyclic()`, `is_cpdag()`, and more.\n- Graph manipulations, such as `add_edge()`, `remove_node()`, and more.\n- Graph metrics, such as `shd()` and `aid()`.\n\n## How it works\n\n`caugi` graphs are represented in a compact Compressed Sparse Row (CSR)\nformat in Rust. `caugi` works with a front-loading philosophy. Since the\n`caugi` graph is stored in a CSR format, mutations of the graph is\ncomputationally expensive compared to other graph storage systems, *but*\nit allows for very fast querying. Additionally to the storage format of\nthe graph itself, `caugi` also stores additional information about node\nrelations in such a way that it allows for faster queries without\nblowing up the object too much.\n\nTo accommodate for the cost of mutations, `caugi` graphs are built\nlazily. This means that when you mutate the graph, for example by adding\nedges to it, the graph edits are stored in R, but not in Rust. When you\nthen need to query the graphs, the graph will rebuild itself in Rust,\nand the query will be executed on the newly built graph. You can also\nuse the `build(cg)` function to force building the graph in Rust at any\ntime.\n\n## Why?\n\nIt’s fast, *dawg*

Description

Usage

Arguments

Value

See Also

Examples