7.3.2.2. Transfer Function

  • Setup a Transfer Function for Payoffs

    Setting up a transfer function for payoffs in a game in MLPro-GT-Native is very simple. There are four main steps, as following:

    from mlpro.gt.native.basics import *
from mlpro.bf.physics.basics import *

class MyTransferFunction(TransferFunction):

    def _set_function_parameters(self, p_args) -> bool:
        # Step 1.1
        ...

    def _custom_function(self, p_input, p_range=None):
        # Step 1.2 (Optional)
        ...

class MyPayoff(GTFunction):

    def _setup_transfer_functions(self):
        # Step 2
        TF = MyTransferFunction(...)

        self._add_transfer_function(p_idx=0, p_transfer_fct=TF)
        self._add_transfer_function(p_idx=1, p_transfer_fct=TF)
        ...

class MyPayoffMatrix(GTPayoffMatrix):

    def _call_mapping(self, p_input:str, p_strategies:GTStrategy) -> float:
        # Step 3.1
        ...

    def _call_best_response(self, p_element_id:str) -> float:
        # Step 3.2
        ...

    def _call_zero_sum(self) -> bool:
        # Step 3.3
        ...


class MyGame(GTGame):

    C_NAME  = 'MyGame'

    def _setup(self, p_mode, p_ada:bool, p_visualize:bool, p_logging) -> Model:

        ...

        # Step 4
        self._payoff = MyPayoffMatrix(
            p_function=MyPayoff(
                p_func_type=GTFunction.C_FUNC_TRANSFER_FCTS
                ),
            p_player_ids=coal_ids
        )

        ...

  • Prerequisites