dairlib/bindings/pydairlib/analysis/tray_balance_study/parameter_study_analysis.py at 215e0b7d05ce09feaf8a89a745fccb84c95f1f1c · DAIRLab/dairlib · 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
import subprocess
import time
from tqdm import *
import lcm
from bindings.pydairlib.common import plot_styler, plotting_utils
import pydairlib.analysis.mbp_plotting_utils as mbp_plots
from bindings.pydairlib.lcm.process_lcm_log import get_log_data
import matplotlib.pyplot as plt
import numpy as np
import yaml
import dairlib


def process_c3_channel(data):
    t = []
    x = []
    for msg in data:
        t.append(msg.utime / 1e6)
        x.append(msg.state)
    t = np.array(t)
    x = np.array(x)

    return {'t': t,
            'x': x}

def load_c3_channels(data, c3_target_state_channel, c3_actual_state_channel, c3_debug_output_channel):
    c3_target = process_c3_channel(data[c3_target_state_channel])
    c3_actual = process_c3_channel(data[c3_actual_state_channel])
    c3_output = mbp_plots.process_c3_debug(data[c3_debug_output_channel])
    return c3_target, c3_actual, c3_output

def load_lcm_logs():
    config_file = 'parameter_study_config.yaml'
    parameters_directory = 'examples/franka/parameters/'
    lcm_channels_sim = 'lcm_channels_simulation.yaml'

    with open(parameters_directory + lcm_channels_sim) as f:
        filedata = f.read()
        lcm_channels = yaml.load(filedata)

    with open('bindings/pydairlib/analysis/tray_balance_study/' + config_file) as f:
        filedata = f.read()
        param_study = yaml.load(filedata)

    start_time = 0
    duration = -1

    c3_channels = {lcm_channels['c3_target_state_channel']: dairlib.lcmt_c3_state,
                   lcm_channels['c3_actual_state_channel']: dairlib.lcmt_c3_state,
                   lcm_channels['c3_debug_output_channel']: dairlib.lcmt_c3_output}
    callback = load_c3_channels
    mass_range = np.arange(0.5, 2.0, 0.05)
    # mu_range = np.arange(0.3, 0.8, 0.01)
    effective_mu_range = np.arange(0.1, 0.71, 0.1)

    print(effective_mu_range.shape[0])
    all_successes = np.zeros((effective_mu_range.shape[0], 3))
    for i in range(effective_mu_range.shape[0]):
        print('%02d', effective_mu_range[i])
        # print(mass_range[i])
        # ps = plot_styler.PlotStyler(nrows=2)

        successes = np.zeros(3)

        for j in range(5):
            # ps = plot_styler.PlotStyler()
            log_filename = param_study['results_folder'] + param_study['parameter'][2] + '/simlog-' + '%02d_%1d' % (i, j)
            log = lcm.EventLog(log_filename, "r")
            c3_target, c3_actual, c3_output = \
                get_log_data(log, c3_channels, start_time, duration, callback, False,
                             lcm_channels['c3_target_state_channel'], lcm_channels['c3_actual_state_channel'],
                             lcm_channels['c3_debug_output_channel'])
            length = min(c3_target['t'].shape[0], c3_actual['t'].shape[0])
            # ps.plot(c3_actual['t'], c3_actual['x'][:, 8:9], subplot_index=0)
            # ps.plot(c3_actual['t'], c3_actual['x'][:, 9:10], subplot_index=0)
            # ps.plot(c3_actual['t'], c3_actual['x'][:, 1:2], subplot_index=0)
            # ps.plot(c3_actual['t'], c3_actual['x'][:, 2:3], subplot_index=0)
            first_target = np.array([0.45, 0, 0.485])
            second_target = np.array([0.45, 0, 0.6])
            third_target = np.array([0.7, 0.0, 0.485])
            reached_first_target = np.any(np.all(np.isclose(c3_target['x'][:, 7:10], second_target, atol=1e-3), axis=1))
            reached_second_target = np.any(np.all(np.isclose(c3_target['x'][:, 7:10], third_target, atol=1e-3), axis=1))
            reached_third_target = reached_second_target and np.linalg.norm(c3_target['x'][:length, 7:10] - c3_actual['x'][:length, 7:10], axis=1)[-1] < 0.1
            task_success = np.array([reached_first_target, reached_second_target, reached_third_target])
            # print('reached_targets: ', task_success)
            successes += task_success
            # print('reached_second_target: ', reached_second_target)
            # print('reached_third_target: ', reached_third_target)
            t_c3_slice = slice(c3_output['t'].size)
            # mbp_plots.plot_c3_inputs(c3_output, t_c3_slice, 1, ps)
            # plt.show()
        print(successes)
        all_successes[i] = successes
    # np.save('target_successes_23', all_successes )
    np.save('target_successes_03_26', all_successes )

def plot_logs():
    all_successes = np.load('target_successes_03_25.npy')
    bar_width = 0.025
    bar_positions = np.arange(all_successes.shape[0])
    n = all_successes.shape[0]
    # mu_range = np.arange(0.3, 0.74, 0.05)
    effective_mu_range = np.arange(0.1, 0.71, 0.1)

    plt.rc('legend', fontsize=36)
    plt.rc('axes', labelsize=36, titlesize=36)
    plt.rc('xtick', labelsize=36)
    plt.rc('ytick', labelsize=36)
    # Plotting bars for each task
    for i in range(n):
        # for j in range(3):
        # plt.bar(bar_positions[i], all_successes[i, 2] / 30, width=bar_width)
        # plt.bar(effective_mu_range[i], all_successes[i, 2] / all_successes[i, 0], width=bar_width, color='grey')
        plt.bar(effective_mu_range[i], all_successes[i, 0] / 5, width=bar_width, color='grey')
    plt.ylabel('Target Success Rate')
    plt.xlabel('Tray/End Effector Friction Coefficient')

    # plt.savefig('/home/yangwill/Pictures/plot_styler/' + 'target_1_success')
    # plt.legend()
    # for i in range(3):
    #     plt.bar(bar_positions + i * (n + 1), all_successes[:, i], label=f'Task {i + 1}')

    plt.show()

def main():
    load_lcm_logs()
    # plot_logs()


if __name__ == '__main__':
    main()