phi=phi,
gamma=gamma)
lstd.name = r"FPKF({}) $\alpha$={} $\beta={}$".format(lam, alpha, beta)
lstd.color = "g"
lstd.ls = "-."
methods.append(lstd)
alpha = .5
rg = td.ResidualGradientDS(alpha=alpha, phi=phi, gamma=gamma)
rg.name = r"RG DS $\alpha$={}".format(alpha)
rg.color = "brown"
rg.ls = "--"
methods.append(rg)
alpha = .5
rg = td.ResidualGradient(alpha=alpha, phi=phi, gamma=gamma)
rg.name = r"RG $\alpha$={}".format(alpha)
rg.color = "brown"
methods.append(rg)
brm = td.RecursiveBRMDS(phi=phi, eps=1e5)
brm.name = "BRMDS"
brm.color = "b"
brm.ls = "--"
methods.append(brm)
brm = td.RecursiveBRM(phi=phi, eps=1e5)
brm.name = "BRM"
brm.color = "b"
methods.append(brm)
brm.ls = "--"
methods.append(brm)
brm = td.RecursiveBRM(phi=phi)
brm.name = "BRM"
brm.color = "b"
methods.append(brm)
alpha = 0.5
rg = td.ResidualGradientDS(alpha=alpha, phi=phi)
rg.name = r"RG DS $\alpha$={}".format(alpha)
rg.ls = "--"
methods.append(rg)
alpha = 0.5
rg = td.ResidualGradient(alpha=alpha, phi=phi)
rg.name = r"RG $\alpha$={}".format(alpha)
methods.append(rg)
eta = 0.001
reward_noise = 0.001
P_init = 1.
ktd = td.KTD(phi=phi, gamma=1., P_init=P_init, theta_noise=None, eta=eta,
reward_noise=reward_noise)
ktd.name = r"KTD $\eta$={}, $\sigma^2$={} $P_0$={}".format(
eta, reward_noise, P_init)
#methods.append(ktd)
sigma = 1.
gptdp = td.GPTDP(phi=phi, sigma=sigma)
gptdp.name = r"GPTDP $\sigma$={}".format(sigma)