GitHub

Installation

Requirements

  • Julia 1.11
  • 8 GB RAM recommended (4 GB minimum)

Install (Package not yet registered)

Install directly from GitHub:

using Pkg
Pkg.add(url="https://github.com/atantos/TextAssociations.jl")

When the package gets registered, you’ll be able to do: Pkg.add("TextAssociations").

Quick check

Once installation completes successfully, you can run a basic smoke test to confirm that the package loads and its core functions work correctly:

using TextAssociations

# Basic smoke test from raw text
text = "The quick brown fox jumps over the lazy dog."
df = assoc_score(PMI, text, "the"; windowsize=3, minfreq=1)
@show first(df, min(5, nrow(df)))

This snippet verifies that:

  • The package precompiles without errors.
  • The assoc_score API works directly from raw text input.
  • The output is a DataFrame of collocates with PMI scores.

Indicative corpus analysis workflow:

Once you’ve confirmed the installation, try a minimal corpus-level workflow. This example shows how to load a small dataset, analyze one node word, and export the results:

using TextAssociations

corpus = read_corpus("path/to/data.csv"; text_column=:text)  # or a folder of .txt files
res = analyze_node(corpus, "important", PMI; windowsize=5, minfreq=5)

CSV.write("path/to/output_file.csv", res)

This verifies that:

Corpus loading (read_corpus) and normalization pipelines run correctly. Association metrics (here, PMI) execute over corpus-level contingency tables.

Run the test suite

To run all automated package tests:

using Pkg
Pkg.test("TextAssociations")

Environments (recommended)

Using a dedicated Julia environment is the best way to manage dependencies for your projects. It keeps the package versions used by TextAssociations.jl isolated from those of other projects and makes your setup fully reproducible on another machine or by collaborators.

You can create and activate a clean environment as follows:

On the system terminal, first create a new folder and move into it. This will serve as your project directory:

mkdir("TextAssocDemo") # creates a new folder that will host your project
cd("TextAssocDemo") # changes the working directory into that folder

These commands simply prepare an empty workspace where Julia will store the project’s Project.toml and Manifest.toml files.

Then, on the Julia REPL, activate that directory as your working environment and install the package:

using Pkg; Pkg.activate(".")
Pkg.add(url="https://github.com/atantos/TextAssociations.jl")
# later: Pkg.instantiate()  # reproduce

This setup:

  • Prevents version conflicts with other Julia packages you may use.
  • Lets you reproduce your exact software environment via Project.toml and Manifest.toml.
  • Simplifies sharing your workflow with others (or with future you).

Next steps