Welcome to Grid2Game documentation!

Grid2Game is a python package that allows you to interact with a powergrid (using grid2op action types).

It allows to:

• “play” the game as if you were operating a powergrid

• inspect the state of powergrid visually

• inspect the past state of the grid (with the productions, loads as well as maximum flows on the powerlines)

• design some basic collaboration between an agent (called “assistant” in this package) and a human, by having the assistant proposing actions to the human

• save the current experiment in a standardized manner (using the grid2op runner) that can be read back with grid2op (for example if you want to perform some kind of supervized learning or an agent) or with grid2viz (for manual a posteriori examination)

Quickstart

• Getting started
  • Install grid2game
  • Main Possibilities
  • Start Using grid2game
• First usage example: play the “grid2op game”
  • Choose a scenario
  • Play interactively
  • Choose an action
• Second usage example: Help your agent ? Help from your agent ?
• Third usage example: improve your agent by teaching it

Technical Documentation (work in progress)

This is a work in progress at the moment

Technical Documentation

• How to load an assisstant
• Save experiment
• How to select a scenario or seed
• Step, back, +XXX and “end”
• Control physical units
• The timeline and the “time” control
• Display of the powergrid
• Perform action on the powergrid
• Past temporal data

Indices and tables

• Index

• Module Index

• Search Page