Skip to content

Latest commit

 

History

History
43 lines (32 loc) · 3.57 KB

README.md

File metadata and controls

43 lines (32 loc) · 3.57 KB
Raw

Puppitor

A Python game interface, animation, and AI library designed to allow players to act out a character using physicality rather than dialogue

Please note that this repository is currently being used to maintain a copy of Puppitor separate from its game development branch and as such will not be updated frequently. The system is still in development and will not have documentation or examples created for public use for some time.

Directory Structure

src:

  • core contains the modules needed for basic functionality in Puppitor:

    • affecter.py (contains state tracking and rules application code housed in the Affecter class as well as make_affect_vector() which is used to format a python dictionary for use with Affecter features)
    • action_key_map.py (contains input mapping and state tracking in the Action_Key_Map class for use either with keyboard, and mouse button, input or AI behavior algorithms, can be used to drive changes in state to be applied to affect_vectors using rules stored in an Affecter)

  • secondarycontains potentially useful modules in either driving or reacting to state changes in Puppitor's core modules:

    • animation_structure.py (contains the Animation_Structure class designed for reactive sprite animation based on the states tracked by an Affecter and Action_Key_Map)
    • npc_greedy.py (contains the think function that performs a greedy search over a given affect_vector, the current_emotional_state argument, using a given action_key_map and affecter as the basis for actions to try and effects to evaluate)
    • npc_uct.py (contains the uct_think function that performs Monte Carlo Tree Search based on given Puppitor inputs as well as the Node class for building trees)
    • npc_a_star.py (contains the npc_a_star_think function that performs A* search from a given affect_vector towards a goal_emotion)

tools:

  • prfeu.py (tests a given rule file and affect vector to see if a particular affect is expressable using a selected search algorithm)

    • usage: $ py prfeu.py ./affect_rules/test_passions_rules.json ./key_map.json ./affect_vector.json fear resting neutral 180 0

  • a_star_unit_test.py (tests a given rule file over each of its possible affects by starting an A* search from the least ideal state possible, the goal_emotion value at 0 and every other value in an affect vector at 1. Prints the results of each trial)

    • usage: $ py a_star_unit_test.py ./affect_rules/test_passions_rules.json ./key_map.json resting neutral 180 F

Usage

To use Puppitor simply put affecter.py and action_key_map.py files in the desired directory and have import affecter and import action_key_map lines your project. Note that to use the components of affecter.py you will need to have JSON files formatted as Puppitor rules, examples can be found in affect_rules. Detailed API descriptions to come.

To run the CLI utilities found in tools you will need to move them into a directory containing affecter.py, action_key_map.py, npc_a_star.py, npc_uct.py, and npc_greedy.py as well as a json rule file, json key map file and for prfeu.py a json file with an affect vector in it. Example files can be found in the tools directory.

Example Command Line calls:

for prfeu.py:

$ py prfeu.py ./affect_rules/test_passions_rules.json ./key_map.json ./affect_vector.json fear resting neutral 180 0 -q 2000

for a_star_unit_test.py:

$ py a_star_unit_test.py ./affect_rules/test_passions_rules.json ./key_map.json resting neutral 180 F <optional queue size limit argument>