Creating Formal Contexts in fcaR

In fcaR: Formal Concept Analysis

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(fcaR)

The starting point of any Formal Concept Analysis (FCA) workflow is the Formal Context. A formal context is a triple $K = (G, M, I)$, where $G$ is a set of objects, $M$ is a set of attributes, and $I \subseteq G \times M$ is a binary relation between them.

In fcaR, formal contexts are managed by the FormalContext R6 class. This vignette demonstrates the multiple ways to create or import a FormalContext object.

1. Creating from R Data Structures

The most direct way to create a formal context is by passing a binary matrix (0/1 or TRUE/FALSE) or a data frame to the constructor.

From a matrix

You can manually define a matrix where rows represent objects and columns represent attributes.

# Create a binary matrix
M <- matrix(c(1, 0, 1,
              1, 1, 0,
              0, 1, 1), 
            nrow = 3, 
            byrow = TRUE)

# Assign row and column names (Objects and Attributes)
rownames(M) <- c("Object1", "Object2", "Object3")
colnames(M) <- c("Attribute1", "Attribute2", "Attribute3")

# Create the FormalContext object
fc <- FormalContext$new(M)

# Print the context
fc

From a data frame

You can also use a data.frame. fcaR will attempt to coerce it into a binary matrix. This is useful if your data is already loaded in R from another source.

df <- data.frame(
  has_wings = c(TRUE, TRUE, FALSE),
  can_fly   = c(TRUE, FALSE, FALSE),
  has_legs  = c(TRUE, TRUE, TRUE),
  row.names = c("Eagle", "Penguin", "Dog")
)

fc_df <- FormalContext$new(df)
fc_df

2. Importing from Local Files

fcaR supports several standard file formats used in the FCA community. The FormalContext$new() constructor automatically detects the format based on the file extension.

CXT Format (.cxt)

The Burmeister format (.cxt) is the standard for many classic FCA tools like ConExp.

# We use an example file included in the package
cxt_file <- system.file("contexts", "lives_in_water.cxt", package = "fcaR")

# Load the context
fc_cxt <- FormalContext$new(cxt_file)
fc_cxt

CSV Files (.csv)

You can load a context from a Comma-Separated Values file. It assumes the first column contains object names and the header contains attribute names.

# We use an example file included in the package
csv_file <- system.file("contexts", "airlines.csv", package = "fcaR")

# Load from CSV
# Note: Ensure your CSV contains binary data
fc_csv <- FormalContext$new(csv_file)

# Inspect dimensions
dim(fc_csv)

3. Fetching from the FCA Repository

fcaR includes a powerful feature to download and load curated datasets directly from the online FCA Repository (fcarepository.org).

Browsing the Repository

You don't need to know the filenames by heart. You can browse the available contexts and see their metadata (dimensions, description, source) directly from the console.

# Get the list of available contexts
meta <- get_fcarepository_contexts()

# Print a detailed summary to the console
# (Shows Title, Dimensions, and Description for each entry)
print_repo_details(meta)

Downloading a Specific Context

Once you have identified a context of interest (e.g., "planets_en.cxt" or "animals_en.cxt"), you can download it using fetch_context(). This function handles the download, parsing, and error checking for you.

# Download and load the 'Planets' context
fc_planets <- fetch_context("planets_en.cxt")

# The object is ready for analysis
fc_planets$find_concepts()
fc_planets$concepts$size()

4. Interactive Use: The RStudio Addin

For a more visual and user-friendly experience, fcaR includes an RStudio Addin. This tool provides a Graphical User Interface (GUI) to explore the repository without writing code initially.

How to Launch It

You have two options:

1. Via menu: In RStudio, go to the toolbar, click the "Addins" dropdown menu, and select "Browse FCA Repository" under the fcaR section.
2. Via console: Run the following command:

select_repository_context_addin()

What Does the Addin Do?

1. It opens a floating window with an interactive table of all available contexts in the repository.
2. You can filter and search contexts by filename, title, or description.
3. When you select a row and click "Done", the Addin automatically inserts the necessary code into your console to download that specific context (using fetch_context).

This is the recommended way to discover new datasets and start working quickly.

Summary of Methods

| Source | Function / Method | Description | | :--- | :--- | :--- | | R Matrix / Data Frame | FormalContext$new(x) | Direct creation from in-memory objects. | | Local File (.cxt) | FormalContext$new("file.cxt") | Loads Burmeister format files. | | Local File (.csv) | FormalContext$new("file.csv") | Loads CSV files (expects binary data). | | FCA Repository (Code) | fetch_context("name.cxt") | Downloads and loads from the web. | | FCA Repository (GUI) | select_repository_context_addin() | Graphical interface to search and load. |

fcaR documentation built on July 27, 2026, 5:06 p.m.