PadePSD_python/main.py at main · ankitgupta83/PadePSD_python · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
import sys
import ReactionNetworkExamples as rxn_examples
import numpy as np
import PSD_Estimator as psd
import json
import pickle
import time
import os
import PadePSD
import plotting as plt_file
import matplotlib.pyplot as plt

def main(argv):
    print(argv)
    config_filename = "./Configs/" + argv[1]
    print("The configuration file is: " + config_filename)
    config_file = open(config_filename, )
    config_data = json.load(config_file)
    config_file.close()
    # get the network
    network = getattr(rxn_examples, config_data['network_config']['network_name'])()
    # set up output folder
    results_folder_path = "./Results/" + config_data['network_config']['output_folder_name'] + "/"
    if not os.path.exists(results_folder_path):
        os.makedirs(results_folder_path)

    stop_time = config_data['network_config']['stop_time']
    cut_off_time = config_data['network_config']['cut_off_time']
    output_species = network.species_labels.index(network.output_species_labels[0])

    # generate sampled trajectories if needed
    if config_data['data sampling_config']['new_data_required'] == "True":
        start_time = time.time()
        print("Starting simulated data generation")
        num_time_samples = config_data['data sampling_config']['num_time_samples']
        num_trajectories = config_data['data sampling_config']['num_trajectories']
        sampling_times, states_array = network.generate_sampled_ssa_trajectories(cut_off_time, stop_time,
                                                                                 num_time_samples, num_trajectories)
        print("Finished simulated data generation in %3u seconds" % (time.time() - start_time))
        output_states_array = states_array[:, :, output_species]
        header_list = 'Time'
        for i in range(1, num_trajectories + 1):
            header_list += ',Trajectory {}'.format(i)
        np.savetxt('{}Trajectory_Data.csv'.format(results_folder_path),
                   np.transpose(np.vstack((sampling_times, states_array[:, :, output_species]))),
                   fmt=['%.5f'] + ['%d' for _ in range(num_trajectories)],
                   header=header_list,
                   delimiter=",",
                   comments='')
    else:
        trajectory_data = np.transpose(np.loadtxt('{}Trajectory_Data.csv'.format(results_folder_path),
                                                  delimiter=",", skiprows=1))
        sampling_times = trajectory_data[0, :]
        output_states_array = trajectory_data[1:np.shape(trajectory_data)[0], :]
        del trajectory_data
    # compute PSD from simulated data
    omega_dft, psd_dft, Mean_output_dft, Standard_dev_output_dft = psd.PSD_Estimator_Numerical(sampling_times, output_states_array)
    psd_dft_mean = np.mean(psd_dft, axis=0)
    psd_dft_std = np.std(psd_dft, axis=0)

    s_vals = config_data['Pade_Approximation_config']['s_vals']
    if s_vals[-1] == "infinity":
        s_vals[-1] = np.infty
    order = config_data['Pade_Approximation_config']['order']
    test_s_values = np.array(config_data['Pade_Approximation_config']['test_s_values'])
    p = config_data['Pade_Approximation_config']['rational_approx_degree']
    known_den_poly = config_data['Pade_Approximation_config']['known_den_poly_coeffs']
    # known_num_poly = config_data['Pade_Approximation_config']['known_num_poly_coeffs']
    # estimate Pade derivatives if needed
    if config_data['Pade_Approximation_config']['pade_derivative_estimation_required'] == "True":

        num_trajectories = config_data['Pade_Approximation_config']['num_trajectories']
        num_time_samples = config_data['Pade_Approximation_config']['num_time_samples']
        # set up the Pade PSD object
        start_time = time.time()
        pade_psd = PadePSD.PadePSD(network, stop_time=stop_time, cut_off_time=cut_off_time, s_vals=s_vals,
                                   test_s_values=test_s_values)
        print("Starting estimation of the required Pade derivatives and the direct G(s) estimates")
        pade_dict = pade_psd.EstimatePadeDerivatives(order=order, num_time_samples=num_time_samples, num_trajectories=num_trajectories)
        print("Estimation of the required quantities completed in %3u seconds\n" % (time.time() - start_time))
        # save the data
        Pade_derivative_data_file = open(results_folder_path + "Pade_derivative_data" + ".pkl", "wb")
        pickle.dump(pade_dict, Pade_derivative_data_file)
        Pade_derivative_data_file.close()
    else:
        Pade_derivative_data_file = open(results_folder_path + "Pade_derivative_data" + ".pkl", "rb")
        pade_dict = pickle.load(Pade_derivative_data_file)
    print("\nEstimated Pade derivatives are as follows:\n")
    for i in range(np.shape(s_vals)[0]):
        if s_vals[i] != np.infty:
            print("for s = %3f" % s_vals[i])
        else:
            print("for s = infinity")
        print('order, mean, std.')
        np.savetxt(sys.stdout, np.transpose(
            np.vstack((np.arange(order[i]), pade_dict[s_vals[i]][0][:order[i]], pade_dict[s_vals[i]][1][:order[i]]))),
                   fmt="%d %.3f %.3f")
        print()

    print("Estimated validation test values are as follows:\n")
    print('s mean, std.')
    np.savetxt(sys.stdout,
               np.transpose(np.vstack((test_s_values, pade_dict['Test_s_vals'][0], pade_dict['Test_s_vals'][1]))),
               fmt="%.3f %.3f %.3f")
    total_power = pade_dict['total_power']
    print()


    pade_derivatives = pade_dict
    pade_psd_function, G_p = PadePSD.ConstructMultipointPadeApproximant(pade_derivatives, s_vals, order, p,
                                                                        known_den_poly)
    print("G_p(x):\n", G_p)
    print("PSD Expression:\n", pade_psd_function)
    # save the G_p and PSD
    Gp_data_file = open(results_folder_path + "G_p" + ".pkl", "wb")
    pickle.dump(G_p, Gp_data_file)
    Pade_derivative_data_file.close()

    omega_limit = config_data['Plotting_config']['ratio_omega_max_vs_pi'] * np.pi
    omega_pade, psd_pade = PadePSD.ConstructPadeApproximationPSD(pade_psd_function, omega_limit)

    # compute analytical PSD if available
    if config_data['network_config']['analytical_psd_available'] == "True":
        omega_anal, psd_anal = network.PSDArray(omega_limit)
    else:
        omega_anal = None
        psd_anal = None

    # make a dictionary of all PSD estimations
    psd_dict = {'DFT': [omega_dft, psd_dft_mean, psd_dft_std],
                'Pade': [omega_pade, psd_pade],
                'Analytical': [omega_anal, psd_anal]
                }

    if config_data['Plotting_config']['normalization'] == "True":
        TotalPowerPade = total_power * np.pi
        TotalPowerExperimental = sum(psd_dft_mean) * (omega_dft[1] - omega_dft[0])
        psd_dict['DFT'][1] = psd_dict['DFT'][1] / TotalPowerExperimental
        psd_dict['DFT'][2] = psd_dict['DFT'][2] / TotalPowerExperimental
        psd_dict['Pade'][1] = psd_dict['Pade'][1] / TotalPowerPade
        if config_data['network_config']['analytical_psd_available'] == "True":
            psd_dict['Analytical'][1] = psd_dict['Analytical'][1] / TotalPowerPade

    save_pdf = True
    number_of_plot_trajectories = config_data['Plotting_config']['num_trajectories']
    trajectory_length = config_data['Plotting_config']['trajectory_length']
    plt_file.plotFunctionTrajectory(sampling_times, output_states_array[:number_of_plot_trajectories, :],
                                    number_of_plot_trajectories, trajectory_length, results_folder_path,
                                    save_pdf)
    plt_file.generate_psd_comparison_plot(psd_dict, network.time_unit, results_folder_path, omega_limit, save_pdf)
    plt_file.generate_test_s_values_comparison(test_s_values, pade_dict['Test_s_vals'][0], pade_dict['Test_s_vals'][1],
                                               G_p,
                                               results_folder_path,
                                               save_pdf)
    plt.show()


if __name__ == '__main__':
    main(sys.argv)