6efe8ce53b30db7dd08dd727b23d6acb11796a8a,examples/grid_world.py,,,#,54

Before Change

    rewardQ08 = Parallel(n_jobs=-1)(
        delayed(experiment)(QLearning, .8) for _ in xrange(n_experiment))
    rewardDQ1 = Parallel(n_jobs=-1)(
        delayed(experiment)(DoubleQLearning, 1) for _ in xrange(n_experiment))
    rewardDQ08 = Parallel(n_jobs=-1)(
        delayed(experiment)(DoubleQLearning, .8) for _ in xrange(n_experiment))

After Change

    logger.Logger(3)

    names = {1:"1", 0.8:"08", QLearning:"Q", DoubleQLearning:"DQ"}
    for e in [1, .8]:
        for a in [QLearning, DoubleQLearning]:
            r = Parallel(n_jobs=-1)(
                delayed(experiment)(a, e) for _ in xrange(n_experiment))
            np.save("r" + names[a] + names[e] + ".npy", np.mean(r, axis=0))

Instances