Adventures in Data Science: an RPG built from the data science toolkit

always a work in progress

The core application is functional, but still has a long way to go before I have a game. I'm not guaranteeing that I'll keep it online 24/7 as I'm actively building it out. I'm not even at the point where I'm posting the game link. PM me if you want to see it in action though.

How to play:

Create a world (you'll need to create an account)

The world you create is your own. It's your own saved world with its own history, people and problems. In the creation phase you can tweak the parameters of the world to make the world that you want to play in.

Create a character

Choose the starting characteristics. You will start with the most basic skills and gear; you will have to find out how to get better skills and gear.

Have adventures

• Killing monsters out in the world makes the world safer, and people like you more.
  • plus you might get some cool loot.

Infrastructure

Django

The core app is written using Django. You can clone this app directly into your existing Django project by cloning this repo into your project directory and adding the URLs to your production urls.py view. The tutorial on the Django website should show you how create a Django project if you don't have one.

NOTE that this is a django app that you can add to your django project. Not a stand-alone django project.

Example snippet to add to your urlpatterns in your urls.py:

urlpatterns = [
        path('game', include('game.urls')),
    ]

Also add this to your INSTALLED_APPS in your settings.py

INSTALLED_APPS = [
    'game.apps.gameConfig'
    ]

note that I'm using game as the cloned repository folder, which may be different than the name of this repository. If you cloned your repository under a different name, then you need to change game to your git rep folder name.

You will also need:

• setup for authentication (mine is in the main prodweb django project, not this app) * You couldn't just remove the @login_required decorators as the app uses the user accounts to get the custom world for each user.
• a sql-like database (I'm using Azure SQL even though the app is hosted on EC2), but the default sql-lite should work fine.
  • you will need to migrate your model to the db on your initial setup. See Django docs on doing this.
• Google analytics tags, but that's easy to just remove if you don't want it.

have a look at the conda env for more clues.

Jupyter Notebooks

Notebooks can be served via ssh forwarding. This allows me to test and tinker with the actual deployed code without disrupting the live server. All functions run out of those notebooks are the same as they are in the live app. Notebooks also serve as tests for the logic in the apps. When making changes you can run all the notebooks to ensure that they complete.

You can see the notebooks in the notebooks folder, but my favorites are also posted on my personal blog , which is offline right now due to COVID

Scikit Learn

This is used to:

• Calculate the area of nation states

Pandas and Numpy

• most maps are created in DataFrame objects
• random number and matrix manipulation are done using numpy tools

D3.js

Visuals in notebooks are done using Altair, however the core web UI is all built in D3.js.

generation parameters

The world is generated randomly based on several features that the user decides such as mountain height, average rainfall, number of nation or cities. Then everything is spun up each time a new game begins.

parameters that shape the world

Each of these takes an optional argument params which customizes the world. Default parameters are defined in the __init__ function of each module.

• landscape = contains everything about the natural features of the world.
• culture = contains everything relating to the things built by civilizations such as people and towns.

There are specific modules for nations, towns, and people that all take input from either landscape or culture. All outputs should end up in the world object.

object hierarchy:

world

• culture - json dict of attributes that affect how people behave, used in Person.__init__() functions
• landscape - json dict used in shaping terrain
• nations - list of Nation objects
• towns - list of Town objects
• people - list of Person objects
• df_features - pandas.dataframe of terrain squares
• Character - the player character object, equipment and experience

TODO: Ideas that I have, jot them down quickly so that I can think of them later.

• Time advances in a weekly manner
  • Has seasons
  • Add snow to terrain in wintertime
• Population birthrate is set to sum of danger area surrounding town.