Skip to main content

An agent-based modeling framework for simulating towns of characters in games

Project description


Neighborly

Overview

Neighborly is an extensible, data-driven, agent-based modeling framework designed to simulate towns of characters for games. It is intended to be a tool for exploring simulationist approaches to character-driven emergent narratives. Neighborly's simulation architecture is inspired by roguelikes such as Caves of Qud and Dwarf Fortress.

Currently, Neighborly works best as a narrative data generator. It models characters’ lives, jobs, routines, relationships, and life events. All of these parts are harnessed to produce emergent character backstories as they interact with each other, grow, and change. You can even specify custom characters, businesses, residences, occupations, life events, social rules, and more. Neighborly is meant to be customized to the narrative setting of your creative vision. Check out the samples directory to see how we modeled the popular anime, Demon Slayer.

Neighborly was inspired by lessons learned from working with Talk of the Town and aims to give people better documentation, simpler interfaces, and more opportunities for extension and content authoring.

Core Features

  • Data-driven
  • Add custom character prefabs
  • Add custom business prefabs
  • Define life events and actions to drive narrative generation
  • Define social rules for how characters should feel about each other
  • Define rules for where characters what locations characters should frequent
  • Specify goal-driven behaviors using behavior trees and utility AI
  • Can model various relationship facets like romance, friendship, trust, and respect
  • Collect and export data about agents using Pandas DataFrames
  • Commandline interface (CLI) tool
  • Create plugins to modularize and share custom content
  • Export simulation state to JSON
  • Could be integrated with roguelike development tools like tcod

Not yet supported features

  • Generating characters with a subset of character traits randomly selected from a pool of traits

Installation

Neighborly is available to install from PyPI. This will install the latest official release.

pip install neighborly

If you want to install the most recent changes that have not been uploaded to PyPI, you can install it by cloning the main branch of this repo and installing that.

pip install git+https://github.com/ShiJbey/neighborly.git

Installing for local development

If you wish to download a Neighborly for local development or want to play around with any of the samples, you need to clone or download this repository and install using the editable flag (-e). Please see the instructions below. This will install a Neighborly into the virtual environment along with all its dependencies and a few addition development dependencies such as black and isort for code formatting.

# Step 1: Clone Repository
git clone https://github.com/ShiJbey/neighborly.git

# Step 2a: Create and activate python virtual environment
cd neighborly

# Step 2b: For Linux and MacOS
python3 -m venv venv
source ./venv/bin/activate

# Step 2b: For Windows
python -m venv venv
./venv/Scripts/Activate

# Step 3: Install local build and dependencies
python -m pip install -e ".[development,testing]"

Usage

If you want examples of how to use Neighborly and how to extend it with custom content, please refer to Neighborly's docs and the sample scripts in the samples directory.

Using as a library

Neighborly can be used as a library within a Python script or package. The samples directory contains python scripts that use Neighborly this way. Please refer to them when creating new Plugins and other content.

Writing plugins

Users can extend Neighborly's default content/behavior using plugins. A few default plugins come prepackaged with Neighborly to help people get started. Plugins are implemented as Python packages or modules and are imported by passing their name in the plugins section of the configuration.

Please see the Plugins section of the documentation for more information about authoring plugins.

Running the CLI

Neighborly can be run as a module $ python -m neighborly or commandline $ neighborly script. If you require additional help while running, please use $ python -m neighborly --help or $ neighborly --help.

By default, Neighborly runs a builtin version of Talk of the Town. However, you can configure the simulation settings by creating a neighborlyconfig.yaml file in the same directory where you're running the CLI.

When world generation concludes, Neighborly can write the final simulation data to a JSON file with the seed used for world generation.

Running the Samples

Neighborly provides sample simulations to show users how to customize it to create new story world themes.

# Make sure that you've activated your python virtual environment
# Replace <sample_name>.py with the name of the
# sample you want to run
python ./samples/<sample_name>.py

The samples in the notebooks directory require Jupyter to be installed. So you will need to run the following command to install all the needed dependencies.

python -m pip install -e ".[samples]"

Then start Jupyter and pass the relative path to the notebooks directory. The following assumes that the command is being run from the root of the project.

notebook ./samples/notebooks

Running the Tests

Testing is very important. It is how we are able to ensure that new changes don't break anything. I do my best to keep tests updated, but some tests may be out of date and refer to systems and logic that no longer exist in Neighborly.

Feel free to contribute tests by forking the repo, adding your test(s), and submitting a pull request with a description of your test cases. Your commits should only contain changes to files within the tests directory. If you change any files in other parts of the project, your PR will be rejected.

Please follow the steps below to run Neighborly's test suite. Neighborly uses PyTest to handle unit testing.

# Step 1: Install dependencies for tests
python -m pip install -e ".[testing]"

# Step 2: Run Pytest
pytest

# Step3 : (Optional) Generate a test coverage report
pytest --cov=neighborly tests/

Documentation

The most up-to-date documentation can be found here

Neighborly uses Numpy-style docstrings in code. When adding docstrings for existing or new bits of code please use the following references for how to format your contributions:

Building the documentation

Neighborly's docs are built using Sphinx. Below are instructions for building the docs

# Install the documentation dependencies
python -m pip install -e ".[docs]"

# Build docs as HTML
sphinx-apidoc -o docs/source/module_docs/ src/neighborly
sphinx-build -b html docs/source/ docs/build/html

If you happen to have npm installed, you can use the package.json configuration file to run build, clean build output, and run a test HTTP server.

Contributing

Here are some ways that people can contribute to Neighborly:

  1. Proposing/Implementing new features
  2. Fixing bugs
  3. Providing optimizations
  4. Filing issues
  5. Contributing tutorials and how-to guides
  6. Fixing grammar and spelling
  7. Creating new samples and plugins

If you are interested in contributing to Neighborly, there are multiple ways to get involved, and not all of them require you to be proficient with GitHub. Interested parties can contribute to the core code base of Neighborly and create new content in the way of plugins. I love feedback, and if you have any questions, create a new issue, and I will do my best to answer. If you want to contribute to the core code, free to fork this repository, make your changes, and submit a pull-request with a description of your contribution. Please keep in mind that this project is a tool for creativity and learning. I have a code of conduct to encourage healthy collaboration, and will enforce it if I need to.

Code Style

Neighborly uses Black to handle code style and sorts imports using isort.

DMCA Statement

Upon receipt of a notice alleging copyright infringement, I will take whatever action it deems appropriate within its sole discretion, including removal of the allegedly infringing materials.

The repo image is something fun that I made. I love The Simpsons, and I couldn't think of anyone more neighborly than Ned Flanders. If the copyright owner for The Simpsons would like me to take it down, please contact me. The same takedown policy applies to any code samples inspired by TV shows, movies, and games.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

neighborly-0.11.2.tar.gz (200.9 kB view hashes)

Uploaded Source

Built Distribution

neighborly-0.11.2-py3-none-any.whl (209.8 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page