MCCFN/visualize.py at main · Miraitwo/MCCFN · 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
from functools import reduce
import numpy as np
import pandas as pd
from matplotlib import pyplot as plt
from matplotlib.gridspec import GridSpec
import seaborn as sns
import os


def Draw_Confmat(Confmat_Set, snrs, cfg):
    # 检查 cfg.classes 中的键是否已经是字符串类型，如果不是则进行解码
    if not isinstance(list(cfg.classes.keys())[0], str):
        cfg.classes = {mod.decode('utf-8'): label for mod, label in cfg.classes.items()}

    for i, snr in enumerate(snrs):
        # 创建一个图形对象，并设置画布大小
        fig = plt.figure(figsize=(10, 8))  # 调整宽度和高度，根据需要调整

        # 创建DataFrame
        df_cm = pd.DataFrame(Confmat_Set[i],
                             index=cfg.classes,
                             columns=cfg.classes)

        # 绘制热力图
        heatmap = sns.heatmap(df_cm, annot=True, fmt="d", cmap="Blues")

        # 设置y轴和x轴标签的显示
        heatmap.yaxis.set_ticklabels(
            heatmap.yaxis.get_ticklabels(), rotation=0, ha='right')
        heatmap.xaxis.set_ticklabels(
            heatmap.xaxis.get_ticklabels(), rotation=45, ha='right')

        # 设置坐标轴标签
        plt.ylabel('True label')
        plt.xlabel('Predicted label')

        # 自动调整布局，避免标签被截断
        plt.tight_layout()  # 自动调整布局，防止标签被截断

        # 保存图像
        conf_mat_dir = os.path.join(cfg.result_dir, 'conf_mat')
        os.makedirs(conf_mat_dir, exist_ok=True)
        fig.savefig(conf_mat_dir + '/' + f'ConfMat_{snr}dB.svg', format='svg', dpi=150)

        # 关闭图形
        plt.close()


def Snr_Acc_Plot(Accuracy_list, Confmat_Set, snrs, cfg):
    # 绘制总体准确率
    plt.plot(snrs, Accuracy_list)
    plt.xlabel("Signal to Noise Ratio")
    plt.ylabel("Overall Accuracy")
    plt.title(f"Overall Accuracy on {cfg.dataset} dataset")
    plt.yticks(np.linspace(0, 1, 11))
    plt.grid()

    acc_dir = os.path.join(cfg.result_dir, 'acc')
    os.makedirs(acc_dir, exist_ok=True)
    plt.savefig(acc_dir + '/' + 'acc.svg', format='svg', dpi=150)
    plt.close()

    # 计算每种调制类型的准确率
    Accuracy_Mods = np.zeros((len(snrs), Confmat_Set.shape[-1]))
    for i, snr in enumerate(snrs):
        Accuracy_Mods[i, :] = np.diagonal(Confmat_Set[i]) / Confmat_Set[i].sum(1)

    # 绘制每种调制信号的准确率
    plt.figure(figsize=(10, 8))  # 调整画布大小，增加空间
    mod_labels = [mod.decode('utf-8') if isinstance(mod, bytes) else mod for mod in cfg.classes.keys()]  # 解码字节串为字符串
    for j, mod_label in enumerate(mod_labels):
        plt.plot(snrs, Accuracy_Mods[:, j], label=mod_label)

    plt.xlabel("Signal to Noise Ratio")
    plt.ylabel("Modulation Accuracy")
    plt.title(f"Accuracy by Modulation on {cfg.dataset} dataset")
    plt.grid()

    # 设置图例的位置和防止遮挡
    plt.legend(loc='center left', bbox_to_anchor=(1, 0.5), fontsize=10)  # 右侧显示图例，避免遮挡曲线

    # 保存图像
    plt.savefig(acc_dir + '/' + 'acc_mods.svg', format='svg', dpi=150, bbox_inches='tight')  # 使用 bbox_inches='tight' 确保图像不被裁切
    plt.close()


def save_training_process(train_process, cfg):
    fig1 = plt.figure(figsize=(8, 6))  # Increased figure size for better visibility
    plt.plot(train_process.epoch, train_process.lr_list)
    plt.xlabel("Epoch", fontsize=16)
    plt.ylabel("Learning Rate", fontsize=16)  # Increased font size for better readability
    #plt.title("Learning Rate", fontsize=16)  # Increased font size for better readability
    plt.grid()
    plt.xticks(fontsize=14)  # Set font size for x-axis tick labels
    plt.yticks(fontsize=14)  # Set font size for y-axis tick labels
    plt.tight_layout()  # Adjust layout to fit everything properly
    fig1.savefig(cfg.result_dir + '/' + 'lr.svg', format='svg', dpi=150, bbox_inches='tight')  # Ensure the entire plot is saved
    plt.close()

    fig2 = plt.figure(figsize=(12, 4))
    plt.subplot(1, 2, 1)
    plt.plot(train_process.epoch, train_process.train_loss, "ro-", label="Train loss")
    plt.plot(train_process.epoch, train_process.val_loss, "bs-", label="Val loss")
    plt.legend()
    plt.grid()
    plt.xlabel("Epoch", fontsize=16)
    plt.ylabel("Loss", fontsize=16)  # Increased font size for better readability
    plt.xticks(fontsize=14)  # Set font size for x-axis tick labels
    plt.yticks(fontsize=14)  # Set font size for y-axis tick labels
    plt.subplot(1, 2, 2)
    plt.plot(train_process.epoch, train_process.train_acc, "ro-", label="Train acc")
    plt.plot(train_process.epoch, train_process.val_acc, "bs-", label="Val acc")
    plt.xlabel("Epoch", fontsize=16)
    plt.ylabel("Accuracy", fontsize=16)  # Increased font size for better readability
    plt.legend()
    plt.grid()
    plt.xticks(fontsize=14)  # Set font size for x-axis tick labels
    plt.yticks(fontsize=14)  # Set font size for y-axis tick labels
    plt.tight_layout()  # Adjust layout to fit everything properly
    fig2.savefig(cfg.result_dir + '/' + 'loss_acc.svg', format='svg', dpi=150, bbox_inches='tight')  # Ensure the entire plot is saved
    plt.show()
    plt.close()