如何使用seaborn和查找率创建多个折线图？

Question

我需要使用下面的DataFrame创建多线图的帮助

+---------+
| class   |
+---------+
| English |
+---------+
| Biology |
+---------+
| Computer|
+---------+

数据文件链接data.xlsx或dict数据

        num user_id first_result second_result result        date    point1    point2    point3    point4
0     0   1480R        clear         clear   pass   9/19/2016     clear  consider     clear  consider
1     1    419M     consider      consider   fail   5/18/2016  consider  consider     clear     clear
2     2    416N     consider      consider   fail  11/15/2016  consider  consider  consider  consider
3     3   1913I     consider      consider   fail  11/25/2016  consider  consider  consider     clear
4     4   1938T        clear         clear   pass    8/1/2016     clear  consider     clear     clear
5     5   1530C        clear         clear   pass   6/22/2016     clear     clear  consider     clear
6     6   1075L     consider      consider   fail   9/13/2016  consider  consider     clear  consider
7     7   1466N     consider         clear   fail   6/21/2016  consider     clear     clear  consider
8     8    662V     consider      consider   fail   11/1/2016  consider  consider     clear  consider
9     9   1187Y     consider      consider   fail   9/13/2016  consider  consider     clear     clear
10   10    138T     consider      consider   fail   9/19/2016  consider     clear  consider  consider
11   11   1461Z     consider         clear   fail   7/18/2016  consider  consider     clear  consider
12   12    807N     consider         clear   fail   8/16/2016  consider  consider     clear     clear
13   13    416Y     consider      consider   fail   10/2/2016  consider     clear     clear     clear
14   14    638A     consider         clear   fail   6/21/2016  consider     clear  consider     clear

我需要创建一个条形图和一个折线图，我已经使用data = {'num': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4, 5: 5, 6: 6, 7: 7, 8: 8, 9: 9, 10: 10, 11: 11, 12: 12, 13: 13, 14: 14}, 'user_id': {0: '1480R', 1: '419M', 2: '416N', 3: '1913I', 4: '1938T', 5: '1530C', 6: '1075L', 7: '1466N', 8: '662V', 9: '1187Y', 10: '138T', 11: '1461Z', 12: '807N', 13: '416Y', 14: '638A'}, 'first_result': {0: 'clear', 1: 'consider', 2: 'consider', 3: 'consider', 4: 'clear', 5: 'clear', 6: 'consider', 7: 'consider', 8: 'consider', 9: 'consider', 10: 'consider', 11: 'consider', 12: 'consider', 13: 'consider', 14: 'consider'}, 'second_result': {0: 'clear', 1: 'consider', 2: 'consider', 3: 'consider', 4: 'clear', 5: 'clear', 6: 'consider', 7: 'clear', 8: 'consider', 9: 'consider', 10: 'consider', 11: 'clear', 12: 'clear', 13: 'consider', 14: 'clear'}, 'result': {0: 'pass', 1: 'fail', 2: 'fail', 3: 'fail', 4: 'pass', 5: 'pass', 6: 'fail', 7: 'fail', 8: 'fail', 9: 'fail', 10: 'fail', 11: 'fail', 12: 'fail', 13: 'fail', 14: 'fail'}, 'date': {0: '9/19/2016', 1: '5/18/2016', 2: '11/15/2016', 3: '11/25/2016', 4: '8/1/2016', 5: '6/22/2016', 6: '9/13/2016', 7: '6/21/2016', 8: '11/1/2016', 9: '9/13/2016', 10: '9/19/2016', 11: '7/18/2016', 12: '8/16/2016', 13: '10/2/2016', 14: '6/21/2016'}, 'point1': {0: 'clear', 1: 'consider', 2: 'consider', 3: 'consider', 4: 'clear', 5: 'clear', 6: 'consider', 7: 'consider', 8: 'consider', 9: 'consider', 10: 'consider', 11: 'consider', 12: 'consider', 13: 'consider', 14: 'consider'}, 'point2': {0: 'consider', 1: 'consider', 2: 'consider', 3: 'consider', 4: 'consider', 5: 'clear', 6: 'consider', 7: 'clear', 8: 'consider', 9: 'consider', 10: 'clear', 11: 'consider', 12: 'consider', 13: 'clear', 14: 'clear'}, 'point3': {0: 'clear', 1: 'clear', 2: 'consider', 3: 'consider', 4: 'clear', 5: 'consider', 6: 'clear', 7: 'clear', 8: 'clear', 9: 'clear', 10: 'consider', 11: 'clear', 12: 'clear', 13: 'clear', 14: 'consider'}, 'point4': {0: 'consider', 1: 'clear', 2: 'consider', 3: 'clear', 4: 'clear', 5: 'clear', 6: 'consider', 7: 'consider', 8: 'consider', 9: 'clear', 10: 'consider', 11: 'consider', 12: 'clear', 13: 'clear', 14: 'clear'} } 创建了条形图，其中x =考虑，清除和y =考虑并清除的计数

但是我不知道如何在这种情况下创建折线图

x =日期

y =通过率（％）

合格率是清除/（考虑+清除）的数量

在同一张图中绘制first_result，second_result和结果的速率

该图应如下图

请发表评论或回答我该怎么做。如果我能对日期进行分组并获得比率，那也很好。

Answer 1

这是我的想法：

# first convert all `clear`, `consider` to 1,0
tmp_df = df[['first_result', 'second_result']].apply(lambda x: x.eq('clear').astype(int))

# convert `pass`, `fail` to 1,0
tmp_df['result'] = df.result.eq('pass').astype(int)

# copy the date
tmp_df['date'] = df['date']

# groupby and compute mean, i.e. number_pass/total_count
tmp_df = tmp_df.groupby('date').mean()

tmp_df.plot()

此数据集的输出