deep-learning-based-data-collection/visualize_simulation.py at main · IoTUniurb/deep-learning-based-data-collection · 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
from dataset import Dataset
from utils import printProgressBar
from models import DBPModel, MLModel
import numpy as np
import matplotlib.pyplot as plt
import time


DATASETS_FOLDER = "./dataset-new"
MODELS_FOLDER = "./models-new"
OUT_FOLDER = "./output"
DS = [
    "noweekend/co2_peano_no_weekend.csv",
    # 'noweekend/pm2p5_peano_no_weekend.csv',
    # 'noweekend/rad_peano_no_weekend.csv',
    # 'noweekend/noise_peano_no_weekend.csv'
]

dataset = Dataset(
    DATASETS_FOLDER, DS[0], normalize=True, smooth=True, train_split=3 / 4
)

data = dataset.df_test.values

ml_model = MLModel("model3", 7, 1, dataset, MODELS_FOLDER)
dbp_model = DBPModel(20, 3, 1, 1)

# !mkdir -p "./plots/model3_1_alt/"

# +
ml_buff = data[:20].reshape(1, 20, 1).copy()[:, -7:, :]
dbp_buff = data[:20].reshape(1, 20, 1).copy()
print(ml_buff.shape)
print(dbp_buff.shape)

ml_skip = 0
dbp_skip = 0
ml_counter = 0

for i in range(20, data.shape[0]):
    #     if i > 20:
    #         break
    printProgressBar(i, data.shape[0])

    plt.plot(dataset.inv_scale_data(data[i - 20 : i + 1]), ".--", label="real")
    plt.plot(
        np.arange(13, 20),
        dataset.inv_scale_data(ml_buff.reshape(7, 1)),
        ".-",
        label="ML",
    )
    plt.plot(
        np.arange(0, 20),
        dataset.inv_scale_data(dbp_buff.reshape(20, 1)),
        ".-",
        label="DBP",
    )
    #     print(i)
    real_val = data[i][0]

    ml_pred_val = ml_model.predict(ml_buff)[0][0]
    dbp_pred_val = dbp_model.predict(dbp_buff)[0][0]

    # inverse scale data
    real_val_invscale = dataset.inv_scale_data(real_val)
    ml_pred_val_invscale = dataset.inv_scale_data(ml_pred_val)
    dbp_pred_val_invscale = dataset.inv_scale_data(dbp_pred_val)
    #     print(f'{real_val_invscale     = }')
    #     print(f'{ml_pred_val_invscale  = }')
    #     print(f'{dbp_pred_val_invscale = }')
    # print(f'{real_val_invscale =} {pred_val_invscale =}')

    epsilon = (real_val_invscale * 1) / 100
    # #     print(f"{epsilon=}")

    #     print(f"range [{real_val_invscale-epsilon}, {real_val_invscale+epsilon}]")
    if ml_counter == 0:
        if ml_pred_val_invscale >= (
            real_val_invscale - epsilon
        ) and ml_pred_val_invscale <= (real_val_invscale + epsilon):
            ml_skip += 1
            ml_buff = np.roll(ml_buff, -1)
            ml_buff[:, -1] = ml_pred_val
        else:
            print("ML: send")
            ml_buff = np.roll(ml_buff, -1)
            ml_buff[:, -1] = real_val
            #             for j in range(1,7):
            #                 ml_buff = np.roll(ml_buff,-1)
            #                 ml_buff[:,-1] = data[i+j][0]
            #             i += 7-1
            ml_counter = 7 - 1

    else:
        print("ML: send")
        ml_buff = np.roll(ml_buff, -1)
        ml_buff[:, -1] = real_val

        ml_counter -= 1

    #         ml_buff = np.roll(ml_buff,-1)
    #         p1 = ml_buff[0][0][0]
    #         p2 = real_val
    # #         print(f"{p1=} {p2=}")
    #         omega = (p2-p1)/14
    # #         print(omega)
    #         for j in range(15):
    #             ml_buff[:,j] = p1 + j * omega
    # #         print(ml_buff)
    # #         break
    #         ml_buff[:,-1] = real_val

    if dbp_pred_val_invscale >= (
        real_val_invscale - epsilon
    ) and dbp_pred_val_invscale <= (real_val_invscale + epsilon):
        dbp_skip += 1
        new_val = dbp_pred_val
    else:
        print("DBP: send")
        new_val = real_val
    dbp_buff = np.roll(dbp_buff, -1)
    dbp_buff[:, -1] = new_val

    #     plt.plot(20,real_val,'.',label='real')
    plt.plot(20, ml_pred_val_invscale, ".", label="ML pred")
    plt.plot(20, dbp_pred_val_invscale, ".", label="DBP pred")
    plt.legend()
    plt.title(f"{i = } {ml_skip = } {dbp_skip = }")

    plt.savefig(f"plots/model3_1_alt/{i}.png")
    plt.cla()
# -

print(ml_skip)
print(dbp_skip)