example of policy()? about sorn HOT 3 CLOSED

Thanks for the issue. policy can be any function that maps the state of the agent(sorn) to action parameterized by output_layer_weights. The given example has a stochastic policy. Extra care on next line, using random.choice sampling action from the policy.

def policy(state,w):
    "Implementation of softmax policy"
    z = state.dot(w)
    exp = np.exp(z)
    return exp/np.sum(exp)

from sorn.

I have added the sample policy method in the readme. Thanks again. For further info, please feel free to create a new issue ;)

from sorn.

Thank you for the quick response @Saran-nns !

from sorn.