-
Notifications
You must be signed in to change notification settings - Fork 1
/
analyzeResultsDamping.py
136 lines (121 loc) · 5.96 KB
/
analyzeResultsDamping.py
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
'''
This script computes RMSEs for kinematics and kinetics, and compares the
results across cases. Focus is on studying the influence of the damping
coefficients on the simulations.
'''
# %% Import packages
import os
import numpy as np
from scipy.interpolate import interp1d
from sklearn.metrics import mean_squared_error, r2_score
# %% Settings
cases = ['4', '38', '40', '34', '37']
# %% Fixed settings
pathMain = os.getcwd()
# Load results
pathTrajectories = os.path.join(pathMain, 'Results')
optimaltrajectories = np.load(os.path.join(pathTrajectories,
'optimalTrajectories.npy'),
allow_pickle=True).item()
# Load experimental data
pathData = os.path.join(pathMain, 'OpenSimModel', 'new_model')
experimentalData = np.load(os.path.join(pathData, 'experimentalData.npy'),
allow_pickle=True).item()
subject = 'new_model'
threshold = 5 # vGRF threshold for stance-swing transition
metrics = {}
metrics['RMSE'] = {}
metrics['R2'] = {}
signal_range = {}
# %% Kinematics
jointsToAnalyze = ['knee_angle_r', 'ankle_angle_r']
metrics['RMSE']['kinematics'] = {}
metrics['R2']['kinematics'] = {}
signal_range['kinematics'] = {}
for i, joint in enumerate(jointsToAnalyze):
metrics['RMSE']['kinematics'][joint] = {}
metrics['R2']['kinematics'][joint] = {}
for c, case in enumerate(cases):
c_joints = optimaltrajectories[case]['joints']
c_joint_idx = c_joints.index(joint)
# Reference data
c_ref = experimentalData[subject]["kinematics"]["positions"]["mean"][joint].to_numpy()
c_ref_t = experimentalData[subject]["kinematics"]["positions"]["GC_percent"]
# Find stance-swing transition
c_ref_vGRF = experimentalData[subject]["GRF"]["mean"]['GRF_y_r'].to_numpy()
c_ref_idx_tr = np.argwhere(c_ref_vGRF < threshold)[0][0]
# Select stance phase
c_ref_stance = c_ref[:c_ref_idx_tr]
# Interpolate over 100 data points
c_ref_vec = np.linspace(0, c_ref_stance.shape[0]-1, c_ref_stance.shape[0])
c_ref_vec_N = np.linspace(0, c_ref_stance.shape[0]-1, 100)
set_interp = interp1d(c_ref_vec, c_ref_stance)
c_ref_inter = set_interp(c_ref_vec_N)
signal_range['kinematics'][joint] = np.max(c_ref_inter) - np.min(c_ref_inter)
# Simulated data
c_sim = optimaltrajectories[case]['coordinate_values'][c_joint_idx:c_joint_idx+1, :].flatten()
c_sim_t = optimaltrajectories[case]['GC_percent']
# Find stance-swing transition
c_sim_vGRF = optimaltrajectories[case]['GRF'][1, :].T
c_sim_idx_tr = np.argwhere(c_sim_vGRF < threshold)[0][0]
# Select stance phase
c_sim_stance = c_sim[:c_sim_idx_tr]
# Interpolate over 100 data points
c_sim_vec = np.linspace(0, c_sim_stance.shape[0]-1, c_sim_stance.shape[0])
c_sim_vec_N = np.linspace(0, c_sim_stance.shape[0]-1, 100)
set_interp = interp1d(c_sim_vec, c_sim_stance)
c_sim_inter = set_interp(c_sim_vec_N)
# Compute RMSE
metrics['RMSE']['kinematics'][joint][case] = np.round(mean_squared_error(c_ref_inter, c_sim_inter, squared=False), 2)
metrics['R2']['kinematics'][joint][case] = r2_score(c_ref_inter, c_sim_inter)
# %% Kinetics
metrics['RMSE']['kinetics'] = {}
metrics['R2']['kinetics'] = {}
signal_range['kinetics'] = {}
for i, joint in enumerate(jointsToAnalyze):
metrics['RMSE']['kinetics'][joint] = {}
metrics['R2']['kinetics'][joint] = {}
for c, case in enumerate(cases):
c_joints = optimaltrajectories[case]['joints']
c_joint_idx = c_joints.index(joint)
# Reference data
c_ref = experimentalData[subject]["kinetics"]["mean"][joint].to_numpy()
c_ref_t = experimentalData[subject]["kinetics"]["GC_percent"]
# Find stance-swing transition
c_ref_vGRF = experimentalData[subject]["GRF"]["mean"]['GRF_y_r'].to_numpy()
c_ref_idx_tr = np.argwhere(c_ref_vGRF < threshold)[0][0]
# Select stance phase
c_ref_stance = c_ref[:c_ref_idx_tr]
# Interpolate over 100 data points
c_ref_vec = np.linspace(0, c_ref_stance.shape[0]-1, c_ref_stance.shape[0])
c_ref_vec_N = np.linspace(0, c_ref_stance.shape[0]-1, 100)
set_interp = interp1d(c_ref_vec, c_ref_stance)
c_ref_inter = set_interp(c_ref_vec_N)
signal_range['kinetics'][joint] = np.max(c_ref_inter) - np.min(c_ref_inter)
# Simulated data
c_sim = optimaltrajectories[case]['joint_torques'][c_joint_idx:c_joint_idx+1, :].flatten()
c_sim_t = optimaltrajectories[case]['GC_percent']
# Find stance-swing transition
c_sim_vGRF = optimaltrajectories[case]['GRF'][1, :].T
c_sim_idx_tr = np.argwhere(c_sim_vGRF < threshold)[0][0]
# Select stance phase
c_sim_stance = c_sim[:c_sim_idx_tr]
# Interpolate over 100 data points
c_sim_vec = np.linspace(0, c_sim_stance.shape[0]-1, c_sim_stance.shape[0])
c_sim_vec_N = np.linspace(0, c_sim_stance.shape[0]-1, 100)
set_interp = interp1d(c_sim_vec, c_sim_stance)
c_sim_inter = set_interp(c_sim_vec_N)
# Compute RMSE
metrics['RMSE']['kinetics'][joint][case] = np.round(mean_squared_error(c_ref_inter, c_sim_inter, squared=False), 2)
metrics['R2']['kinetics'][joint][case] = r2_score(c_ref_inter, c_sim_inter)
# %% Percent change as a function of signal range
variables = ['kinematics', 'kinetics']
joints = ['knee_angle_r', 'ankle_angle_r']
changes = {}
for variable in variables:
changes[variable] = {}
for joint in joints:
changes[variable][joint] = {}
for c, case in enumerate(cases[1:]):
RMSE_change = metrics['RMSE'][variable][joint][case] - metrics['RMSE'][variable][joint][cases[c]]
changes[variable][joint][case] = np.round(RMSE_change / signal_range[variable][joint] * 100, 1)