states = [
'hand_px', 'hand_py', 'hand_pz', 'hand_vx', 'hand_vy', 'hand_vz', 'offset'
]
decoding_states = ['hand_vx', 'hand_vz', 'offset']
# TODO transpose beta matrix
truedata = loadmat(
'/Users/sgowda/bmi/workspace/adaptive_ppf/ppf_test_case_matlab_output.mat')
a_ppf = truedata['A'][0, 0]
w_ppf = truedata['W'][0, 0]
A = np.mat(np.diag([a_ppf, a_ppf, 1]))
W = np.mat(np.diag([w_ppf, w_ppf, 0]))
# Instantiate PPF
ppf = ppfdecoder.PointProcessFilter(A, W, beta.T, T_loop)
ppf._init_state()
decoded_output = np.zeros([3, n_iter])
for idx in range(1, n_iter):
ppf(spike_counts[idx, :])
decoded_output[:, idx] = ppf.get_mean()
x_est = truedata['x_est']
plt.figure()
plt.hold(True)
plt.plot(x_est[0, :n_iter], label='matlab')
plt.plot(hand_vel[0, :n_iter], label='handvel')
plt.plot(decoded_output[0, :], label='pyth')
plt.legend()
plt.show()
updater_cont = clda.PPFContinuousBayesianUpdater(decoder_rml)
updater_cont.rho = -1
## RUN
n_iter = X.shape[0]
spike_counts = data['spike_counts']
beta_hist = []
beta_cont_hist = np.zeros([n_iter, n_neurons, 3])
beta_cont_hist[0, :, -1] = beta[:, -1]
I = np.mat(np.eye(3 * n_neurons))
R_diag_neuron = 1e-4 * np.array([0.13, 0.13, 0.06 / 50])
R = np.diag(np.tile(R_diag_neuron, (n_neurons, )))
meta_ppf = ppfdecoder.PointProcessFilter(I, R, np.zeros(3 * n_neurons), dt=dt)
meta_ppf._init_state(init_state=beta_cont_hist[0, :, :].ravel(), init_cov=R)
n_iter = 60000
for n in range(1, n_iter):
if n % batch_size == 0: print(n)
int_kin = np.hstack([np.zeros(3), X[n, 0], 0, X[n, 1], 1])
beta_C = np.array([X[n, 0], X[n, 1], 1])
learner(spike_counts[n - 1, :].reshape(-1, 1), int_kin)
if learner.is_ready():
# calc beta est from batch
int_kin_batch, spike_counts_batch = learner.get_batch()
beta_hist.append(decoder_sb.filt.C)
new_params = updater_sb.calc(int_kin_batch, spike_counts_batch, rho,