# for i in range(num_bidders)] # bidders = [WeberBidder(i, num_rounds, num_bidders, possible_types, type_dist, type_dist_disc) # for i in range(num_bidders)] learner = MDPBidderUAI(num_bidders, num_rounds, num_bidders, possible_types, type_dist, type_dist_disc) learner.learn_auction_parameters(bidders, num_mc) # Plot what the bidder has learned learner.valuations = [.2, .1] learner.calc_expected_rewards() learner.solve_mdp() plot_exp_payments(learner) plot_transition(learner) plot_prob_winning_and_transition(learner) plot_Q_values(learner) plot_price_pdf(learner) print(learner.place_bid(1)) # print(learner.place_bid(2)) # Compare learner to other agents bidders[0].reset() b0 = [0] * len(bidders[0].possible_types) lb0 = [0] * len(learner.possible_types) for t_idx, t in enumerate(learner.possible_types): bidders[0].valuations = [t, t / 2.0] learner.valuations = [t, t / 2.0] if t_idx == 0: learner.calc_expected_rewards() else: learner.calc_terminal_state_rewards() learner.solve_mdp() # b20 = learner.place_bid(2)