本文介绍分析文件变化记录，得出统计信息，并画出文件大小随着时间变化的散点图。

1. 统计信息

文件变化记录格式如下：

2022-05-30 13:58:37,201761000037,reader_writer_pv.c,769
2022-05-30 13:58:37,201821121027,reader_writer_pv.c,769
2022-05-30 13:58:37,202021121001,reader_writer_pv.c,2148
...

分析文件变化记录，得到每位学生的文件变化次数、总变化字节数、速度，核心代码如下：

def calculate_seconds_size_speed(filename):
    # Step 1: accumualte seconds and size
    d_user_seconds_size = {}    # username : [seconds, acculated_size, previous_size]
    with open(filename, 'r', encoding='utf-8') as f:
        for row in csv.reader(f, delimiter=','):
            username = row[1]
            size = int(row[-1])

            if username not in d_user_seconds_size:
                # the first record
                d_user_seconds_size[username] = [0, 0, size]
            else:
                d_user_seconds_size[username] = [d_user_seconds_size[username][0] + 1,
                                                 d_user_seconds_size[username][1] + abs(size - d_user_seconds_size[username][2]),
                                                 size]

    # Step 2: calulate speed and sorted
    d_user_seconds_size_speed = {}
    for username, t in d_user_seconds_size.items():
        if t[0] == 0:
            d_user_seconds_size_speed[username] = [t[0], t[1], 0]
        else:
            d_user_seconds_size_speed[username] = [t[0], t[1], t[1] / t[0]]

    d_user_seconds_size_speed_sorted = OrderedDict(sorted(d_user_seconds_size_speed.items(), key=lambda x: x[1][-1], reverse=True))

    return d_user_seconds_size_speed_sorted

2. 画散点图

指量画出每位学生文件大小随时间变化的散点图，核心代码如下：

def draw_file_changes(input_file, output_dir, d_user_name_info, d_user_seconds_size_speed):
    d_user_time_size = {}
    with open(input_file, 'r', encoding='utf-8') as f: # 2022-05-30 13:58:37,201761000037,reader_writer_pv.c,769

        for row in csv.reader(f, delimiter=','):
            username = row[1]
            size = int(row[-1])
            dt = datetime.datetime.strptime(row[0], '%Y-%m-%d %H:%M:%S')

            # if username != '202021121054': #'202021121018':  #201761000037
            #     continue

            if username not in d_user_time_size:
                d_user_time_size[username] = [(dt, size)]
            else:
                d_user_time_size[username].append((dt, size))

    # draw a graph
    for username, lists in d_user_time_size.items():
        # use user_info as output filename
        if username == '201761000037':
            username_os = 'qiankun'
        else:
            username_os = 'u{}'.format(username)

        user_info = d_user_name_info[username_os]
        output_filename = '{}/{}.jpg'.format(output_dir, user_info)

        # draw a graph
        fig, ax = plt.subplots()

        columns = list(zip(*lists))
        # plt.plot(columns[0], columns[1], 'r-x', linewidth=0.5, label='20202112****') # label=user_info
        plt.plot(columns[0], columns[1], 'r-x', linewidth=0.5, label=user_info)  # label=user_info

        seconds, size, speed = d_user_seconds_size_speed[username]
        title = '文件变化次数={}   累计文件大小={}   速度={}'.format(seconds, size, round(speed, 2))

        plt.title(title)
        plt.xlabel('时间')
        plt.ylabel('文件大小（单位：字节）')
        plt.grid(True)
        plt.legend(loc='best')


        start_dt = datetime.datetime.strptime('2022-05-30 13:50', '%Y-%m-%d %H:%M')
        end_dt = datetime.datetime.strptime('2022-05-30 15:40', '%Y-%m-%d %H:%M')
        plt.xlim(start_dt, end_dt)

        # date format
        date_format = mdates.DateFormatter('%H:%M')
        ax.xaxis.set_major_formatter(date_format)

        #plt.xticks(columns[0]) #rotation='vertical'
        fig.tight_layout()

        plt.savefig(output_filename)

完整的代码见file_changes_analysis.py。