refactor policy to not depend on phase

keithgw / wingspan

A Wingspan AI that uses reinforcement learning to learn to play the wingspan board game.

2 stars 1 forks source link

refactor policy to not depend on phase #65

Open keithgw opened 7 months ago

keithgw commented 7 months ago

Since policies always receive legal actions to choose from, having the code depend on phase is uncessary. Policy should have one method for choosing actions whether they be PAB/GF/DAB, choosing a bird from the hand to play, or choosing a bird to draw.