数据存储与绘制

1.数据输出与存储：

'''
数据输出
'''
print('\n----------------------------OUTPUT----------------------------\n')
stats_list = []

for k, v in results.items():
    mean_value = np.mean(v)
    std_value = np.std(v)
    median_value = np.median(v)
    max_value = np.max(v)
    min_value = np.min(v)
    stats_list.append([k, mean_value, std_value, median_value, max_value, min_value])  # 存储一下，方便保存到csv文件

    print(
        "{:<15} Mean = {:>7.3f}, Std = {:>7.3f}, Median = {:>7.3f}, Max = {:>7.3f}, Min = {:>7.3f}".format(
            k, mean_value, std_value, median_value, max_value, min_value))
'''
数据保存
'''

# N1.  测试结果写入csv文件
headers = ["Metric", "Mean", "Std", "Median", "Max", "Min"]
with open('stats.csv', mode='w', newline='') as file:
    writer = csv.writer(file)
    writer.writerow(headers)  # 写入标题行
    writer.writerows(stats_list)  # 写入统计结果
print("\nData has been written to stats.csv\n")

2.数据读取与绘制

import csv
import matplotlib.pyplot as plt

# 读取CSV文件并存储数据
def read_csv(filename):
    with open(filename, mode='r', newline='') as csvfile:
        reader = csv.DictReader(csvfile)
        data = {rows['Metric']: {
            'Mean': float(rows['Mean']),
            'Std': float(rows['Std']),
            'Median': float(rows['Median']),
            'Max': float(rows['Max']),
            'Min': float(rows['Min'])
        } for rows in reader}
    return data

# 分组数据
def group_data(data):
    similarity = {metric: stats for metric, stats in data.items() if stats['Max'] < 2}
    distance = {metric: stats for metric, stats in data.items() if stats['Max'] >= 2}
    return similarity, distance

# 绘制图表
def plot_data(data_dict, title, filename):
    metrics = list(data_dict.keys())
    means = [data_dict[metric]['Mean'] for metric in metrics]
    stds = [data_dict[metric]['Std'] for metric in metrics]
    medians = [data_dict[metric]['Median'] for metric in metrics]
    max_values = [data_dict[metric]['Max'] for metric in metrics]
    min_values = [data_dict[metric]['Min'] for metric in metrics]

    fig, ax = plt.subplots(figsize=(10, 6))
    bar_width = 0.35
    indices = range(len(metrics))

    bars = plt.bar(indices, means, bar_width, color='skyblue', yerr=stds, capsize=5, label='Mean ± Std')

    # 为每个柱子添加均值标签
    for i, mean_value in enumerate(means):
        plt.text(i+0.2, mean_value + stds[i] * 0.05, f'{mean_value:.3f}', ha='center', va='bottom')

    for i, metric in enumerate(metrics):
        plt.plot(indices[i], medians[i], 'g^', markersize=8)  # 绿色三角表示中位数
        plt.plot(indices[i], max_values[i], 'mv', markersize=8)  # 紫色菱形表示最大值
        plt.plot(indices[i], min_values[i], 'k<', markersize=8)  # 黑色尖角表示最小值

    plt.title(title)
    plt.xlabel('Metrics')
    plt.ylabel('Values')
    plt.xticks(indices, metrics, rotation=45)
    plt.legend()
    plt.tight_layout()
    plt.savefig(filename)
    plt.show()

# 主程序
if __name__ == '__main__':
    filename = 'stats.csv'
    data = read_csv(filename)
    similarity, distance = group_data(data)

    plot_data(similarity, 'Similarity', 'Similarity.png')
    plot_data(distance, 'Distance', 'Distance.png')