ishikota/HoleCardHandicapper

NeuralNet for poker hands evaluation

HoleCardHandicapper

HoleCardHandicapper makes a prediction of heads-up win rates on poker hands.

from holecardhandicapper.model.neuralnet import Neuralnet
from pypokerengine.engine.card import Card

model = Neuralnet("preflop")
model.compile()
holecard = [Card(suit=Card.HEART, rank=1), Card(suit=Card.SPADE, rank=1)]
model.predict(holecard) # returns 0.850103

You can check out sample code for all street (preflop, flop, turn, river)

• https://github.com/ishikota/HoleCardHandicapper/blob/master/sample/how_to_use.py

Installation

pip install HoleCardHandicapper

Dependency

• Keras
  • To make a prediction, HoleCardHandicapper internally uses keras library.
• PyPokerEngine
  • To represents trump card, we use pypokerengine.engine.card.Card class.