interface_scoliose/interpolation.py at main · leadroletroy/interface_scoliose · 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
# Tests and plots to fins the best interpolation splines

import json
import math
import os
import numpy as np
from scipy.interpolate import CubicSpline, interp1d, make_interp_spline, UnivariateSpline, splev, splrep

import matplotlib.pyplot as plt

path = r'D:\StageE23\Data\Ete_2022\Participant06\autocorrection\Prise02'

jsonfile = open(path+'/Positions/positions_corrigees.json')
dict_coordo_labels = json.load(jsonfile)

labels = list(dict_coordo_labels['image1'].keys())

x_axis = range(1, len(list(dict_coordo_labels.keys()))+1)
x, y = {}, {}
cs, spl, s1d, interp = {}, {}, {}, {}
for l in labels:
    y.update({l : [[], []]})
    x.update({l: []})
    cs.update({l: [[], []]})
    spl.update({l: [[], []]})
    s1d.update({l: [[], []]})
    interp.update({l: [[], []]})
for im, coordos in dict_coordo_labels.items():
    for l, c in coordos.items():
        if not np.isnan(c[0]):
            y[l][0].append(c[0])
            y[l][1].append(c[1])
            x[l].append(int(im[5:]))

for l in labels:
    m = len(x[l])
    sm = m-math.sqrt(2*m)
    print(sm)
    std_0 = np.std(y[l][0])
    std_1 = np.std(y[l][1])
    print(std_0, std_1)
    w = np.ones(m) #poids de 1 à tous les points
    w[0] = 5
    w[-1] = 5
    cs[l][0] = CubicSpline(x[l], y[l][0], bc_type='clamped')
    cs[l][1] = CubicSpline(x[l], y[l][1], bc_type='clamped')
    s1d[l][0] = splrep(x[l], y[l][0], w, k=3, s=sm/3)
    s1d[l][1] = splrep(x[l], y[l][1], w, k=3, s=sm/3)
    interp[l][0] = interp1d(x[l], y[l][0], 'cubic')
    interp[l][1] = interp1d(x[l], y[l][1], 'cubic')
    #mss[l][0] = make_smoothing_spline(x[l], y[l][0])

fig, ax = plt.subplots()
for l in labels:
    ax.scatter(x[l], y[l][0], label=f'{l} data')
    ax.plot(x_axis, cs[l][0](x_axis), label=f'{l} cubic')
    ax.plot(x_axis, splev(x_axis, s1d[l][0]), label=f'{l}spl')
    #ax.plot(x_axis, interp[l][0](x_axis), label=f'{l}interp')
ax.legend()
plt.show()

fig, ax = plt.subplots()
for l in labels:
    ax.scatter(x[l], y[l][1], label=f'{l} data')
    ax.plot(x_axis, cs[l][1](x_axis), label=f'{l} cubic')
    ax.plot(x_axis, splev(x_axis, s1d[l][1]), label=f'{l}spl')
    #ax.plot(x_axis, interp[l][1](x_axis), label=f'{l}interp')
ax.legend()
plt.show()

def interpolate_spline(dict_coordo_labels, labels):
    x, y = {}, {}
    splines, spl = {}, {}

    for l in labels:
        y.update({l : [[], []]})
        x.update({l: []})
        splines.update({l: [[], []]})
        spl.update({l: [[], []]})
    for im, coordos in dict_coordo_labels.items():
        for l, c in coordos.items():
            if not math.isnan(c[0]):
                y[l][0].append(c[0])
                y[l][1].append(c[1])
                x[l].append(im[5:])

    for l in labels:
        m = len(x[l])
        sm = m-math.sqrt(2*m)
        w = np.ones(m) #poids de 1 à tous les points
        w[0] = 5
        w[-1] = 5
        cs_x = CubicSpline(x[l], y[l][0])
        cs_y = CubicSpline(x[l], y[l][1])
        splines[l][0] = cs_x
        splines[l][1] = cs_y

    return splines

'''
def smooth_splines(splines):
    splines_smooth = {}
    if m > 5:
        smooth_x = splrep(x[l], y[l][0], w, k=5, s=sm)
        smooth_y = splrep(x[l], y[l][1], w, k=5, s=sm)
        spl[l][0] = smooth_x
        spl[l][1] = smooth_y
        splines_smooth[l][0] = splev(x_axis, spl[l][0])
        splines_smooth[l][1] = splev(x_axis, spl[l][1])
'''

def graph(splines, splines_smooth):
    x_axis = np.arange(len(list(dict_coordo_labels.keys())))
    if not 'interpolate' in os.listdir(path):
        os.mkdir(path+ '/interpolate/')
    fig, (ax1, ax2, ax3) = plt.subplots(1,3)
    ax1.scatter(x_axis, splines['G'][0](x_axis), s=5, label=f'G x')
    ax1.plot(x_axis, splines_smooth['G'][0], label=f'G x smooth')
    ax1.legend()
    ax2.scatter(x_axis, splines['D'][0](x_axis), s=5, label=f'D x')
    ax2.plot(x_axis, splines_smooth['D'][0], label=f'D x smooth')
    ax2.legend()
    for l in ['C', 'T', 'L']:
        ax3.scatter(x_axis, splines[l][0](x_axis), s=5, label=f'{l} x')
        ax3.plot(x_axis, splines_smooth[l][0], label=f'{l} x smooth')
    ax3.legend()
    plt.show()
    plt.close()

    fig, (ax4, ax5, ax6) = plt.subplots(1,3)
    ax4.scatter(x_axis, splines['C'][0](x_axis), s=5, label=f'C x')
    ax4.plot(x_axis, splines_smooth['C'][0], label=f'C Y smooth')
    ax4.legend()
    ax5.scatter(x_axis, splines['L'][0](x_axis), s=5, label=f'L x')
    ax5.plot(x_axis, splines_smooth['L'][0], label=f'L Y smooth')
    ax5.legend()
    for l in ['G', 'D', 'C']:
        ax6.scatter(x_axis, splines[l][0](x_axis), s=5, label=f'{l} Y')
        ax6.plot(x_axis, splines_smooth[l][0], label=f'{l} Y smooth')
    ax6.legend()
    plt.show()
    plt.close()