def experiment(env, step_type, training_rounds, gamma, lr, train_ep, oneshot, test_ep, save_folder, device, config):
"""
Run an experiment where agents are trained and tested (including cross-play)
training_rounds is the number of pairs of policies trained
"""
step_func = step_funcs[step_type]
print("---- Starting ----")
### Train policies
p1s, p2s = train_policies(env, training_rounds, step_func, train_ep, gamma, lr, device)
qs = QuickSaver(subfolder=save_folder)
qs.save_json(config, name='config')
### Save policies
list_p1s = list(map(lambda p: p.tolist(), p1s))
list_p2s = list(map(lambda p: p.tolist(), p2s))
save_results(qs, 'Pols', list_p1s, list_p2s)
for name, prior in env.generate_test_priors():
print("Testing prior", name, "...")
### Test policies
r1s, r2s = several_test(env, prior, p1s, p2s)
xr1s, xr2s = several_cross_test(env, prior, p1s, p2s, n_crosses=training_rounds)
### Plot results
plot_results(env, prior, r1s, r2s, color='orange')
plot_results(env, prior, xr1s, xr2s, color='blue')
save_plot(qs, name + '_results')
### Save results
save_results(qs, name + '_Pfs', r1s, r2s)
save_results(qs, name + '_XPfs', xr1s, xr2s)
def validation(self):
ntest = self.fit_nrows - self.ntrain
steps_test = int(ntest / configs['data']['batch_size'])
tesize = steps_test * configs['data']['batch_size']
data_gen_test = self.dl.generate_clean_data(
configs['data']['filename_clean'],
size=tesize,
batch_size=configs['data']['batch_size'],
start_index=self.ntrain)
print('> Testing model on', ntest, 'data rows with', steps_test,
'steps')
predictions = self.model.predict_generator(
self.generator_strip_xy(data_gen_test), steps=steps_test)
# Save our predictions
with h5py.File(configs['model']['filename_predictions'], 'w') as hf:
dset_p = hf.create_dataset('predictions', data=predictions)
dset_y = hf.create_dataset('true_values', data=self.ture_values)
plot.plot_results(predictions[:800], self.ture_values[:800])
# start_time_1 = time()
# while model_f.t >= model_f.t_final and model_f.condition:
# '''
# Generates the field, which can start at a different redshift from the cluster
# '''
# start_time_2 = time()
# model_f.evolve_field()
# model_f.update_ssfr_params()
# model_f.update_step() # advances t, z to next step
# #print('\t time per step: {:.2f}s'.format(time() - start_time_2))
# #print('~~~~~~~~~~~~~~~~~~~~~~~')
# print('Total ellapsed time for Field: {:.2f}s'.format(time() - start_time_1))
# model_f.parse_masked_mass_field()
# model_c.parse_masked_mass_cluster()
plot_results(params, model_c, model_f)
# total_stel_mass = np.sum(np.power(10,model_c.final_mass_cluster_SF)) + np.sum(np.power(10, model_c.final_mass_cluster_Q))
# total_stel_mass_per_cluster = total_stel_mass/n_clusters
# print('Total stellar mass of cluster: ', np.log10(total_stel_mass_per_cluster))
# p.close()
# p.join()
image_path = f"{args.input_path}/frames/"
detections_path = f"{args.input_path}/{'annotations.mat' if args.annotations else 'observations.mat'}"
mat = scipy.io.loadmat(detections_path)
detections = mat['annotations'] if args.annotations else mat['observations']
file_paths = sorted(glob(f"{image_path}/*.png"))
tracker = Tracker(args.gating_area, args.use_hungarian, args.use_kalman, args.use_mahalanobis)
for i, file_path in enumerate(file_paths):
img = cv2.imread(file_path)
img_tracks = img.copy()
if i > 0:
# previous measurement in red; indexing starts with 1
utils.plot_detections(img, detections[detections[:, 0] == i, 1:], (0, 0, 255))
# current measurement in blue
utils.plot_detections(img, detections[detections[:, 0] == i + 1, 1:], (255, 255, 0))
tracks, debug_info = tracker.update([{'pos': pos} for pos in detections[detections[:, 0] == i+1, 1:]])
if args.use_kalman:
# plot white predicted states
utils.plot_detections(img, np.array(debug_info['predicted_positions']), (255, 255, 255))
utils.plot_tracks(img_tracks, tracks, num_colors=15)
utils.plot_results(img, img_tracks)