https://github.com/HeardACat/castle

You know like...a sand castle in a sandbox or something? Basically a setup to play gym environments
https://github.com/HeardACat/castle

Last synced: 3 months ago
JSON representation

You know like...a sand castle in a sandbox or something? Basically a setup to play gym environments

• Host: GitHub
• URL: https://github.com/HeardACat/castle
• Owner: HeardACat
• License: mit
• Created: 2020-07-29T00:20:22.000Z (almost 5 years ago)
• Default Branch: master
• Last Pushed: 2020-10-11T04:31:04.000Z (over 4 years ago)
• Last Synced: 2025-01-31T17:19:59.302Z (5 months ago)
• Language: Python
• Size: 111 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# castle

You know like...a sand castle in a sandbox or something? Basically a setup to play gym environments

![Castle Example](castle-movement.gif)

![Castle Example](castle.gif)

**Why?**

I like creating gym environments. But I want to play inside them too. It would be pretty cool

to create simple games in pure python with like a couple hundred lines of code; even if it is

an ASCII-esque game.

Rendering games as "rgb_array" is nice if you have all the information to do it, but most of the time I don't have it,

nor do I "care" enough.

**How will the game work?**

My intent is that environments will use several conventions(?) (or is it grammatically better to say "semantics") which a generic wrapper will use. These are:

*  `get_board()` - returns a numpy array of 2 or 3 dimensions. If 2 dimensions then it displays the "height x width", if 3 dimensions then it displays "layers x height x width", especially if items can occupied the same position. 

*  `get_info()` - provides the mapping from the numpy array to the icon which is displayed or the sprite which is used

Types of games to target:

-  Games which run based on a "frame-rate" (e.g. atari-esque)

-  Games which wait for input (e.g. solving a path finding environment)

**Examples**

See `example/` folder. The various scenarios and approaches are shown there:

*  `blocking`: `simple2x2` whereby you navigate a maze

*  `random`: `movement` - a demo example whereby the agent moves around at random

*  `async`: `catch` - a pseudo-async example for an atari-esque game

**Goals and Milestones**

*  Ascii wrapper

*  Ascii wrapper with `colorama`

*  Pyglet with ascii

*  Pyglet with sprites

*  Kivy for cross app development. Need to remove any numpy or gym dependencies to build it better.

Who knows!