如何旋转 pandas 数据框（有两个 headers）

Question

我有一个 Pandas 数据框，如下所示：

Age	USA	USA	USA	UK	UK	UK
Age	male	female	total	male	female	total
2-year-old	2	3	5	1	1	2
3-year-old	8	8	16	7	9	16

事实上我有两行headers（美国+男性；美国+女性；..）

CSV-File (test.csv):

;USA;USA;USA;UK;UK;UK
Age;male;female;total;male;female;total
2-year-old;2;3;5;1;1;2
3-year-old;8;8;16;7;9;16

我的python-code:

df = pd.read_csv('test.csv', 
                 delimiter=";",
                 header=[0,1])

df = df.rename(columns={'Unnamed: 0_level_0': 'Age'})

如何旋转 pandas 数据框以获得以下结果？

Age	Country	Gender	frequency
2-year-old	USA	male	2
2-year-old	USA	female	3
2-year-old	UK	male	1
2-year-old	UK	female	1
3-year-old	USA	male	8
3-year-old	USA	female	8
3-year-old	UK	male	7
3-year-old	UK	female	9

编辑：

开始 Table:

Kode	Country	Procedure	male	male	female	female
Kode	Country	Procedure	two-year-old	three-year-old	two-year-old	three-year-old
1a	US	proc_1	4	6	3	6
1a	UK	proc_1	2	3	5	1
1b	US	proc_2	15	3	5	2
1b	UK	proc_2	8	4	7	3

CSV:

Code;Country;Procedure;male;male;female;female
Code;Country;Procedure;two-year-old;three-year-old;two-year-old;three-year-old
1a;US;proc_1;4;6;3;6
1a;UK;proc_1;2;3;5;1
1b;US;proc_2;15;3;5;2
1b;UK;proc_2;8;4;7;3

结果Table：

Code	Country	Procedure	Gender	Age	Frequency
1a	US	proc_1	male	two-year-old	4
1a	US	proc_1	male	three-year-old	6
1a	US	proc_1	female	two-year-old	3
1a	US	proc_1	female	three-year-old	6
1a	UK	proc_1	male	two-year-old	2
1a	UK	proc_1	male	three-year-old	3
1a	UK	proc_1	female	two-year-old	5
1a	UK	proc_1	female	three-year-old	1
1b	...

Answer 1

使用DataFrame.set_index with DataFrame.stack，如果需要删除总计添加drop:

df = (df.drop('total', axis=1, level=1)
        .set_index(df.columns[0])
        .stack([0,1])
        .rename_axis(['Age','Country','Gender'])
        .reset_index(name='frequency'))
print (df)
          Age Country  Gender  frequency
0  2-year-old      UK  female          1
1  2-year-old      UK    male          1
2  2-year-old     USA  female          3
3  2-year-old     USA    male          2
4  3-year-old      UK  female          9
5  3-year-old      UK    male          7
6  3-year-old     USA  female          8
7  3-year-old     USA    male          8

或者：

df = (df.set_index(df.columns[0])
        .stack([0,1])
        .rename_axis(['Age','Country','Gender'])
        .reset_index(name='frequency'))
print (df)
           Age Country  Gender  frequency
0   2-year-old      UK  female          1
1   2-year-old      UK    male          1
2   2-year-old      UK   total          2
3   2-year-old     USA  female          3
4   2-year-old     USA    male          2
5   2-year-old     USA   total          5
6   3-year-old      UK  female          9
7   3-year-old      UK    male          7
8   3-year-old      UK   total         16
9   3-year-old     USA  female          8
10  3-year-old     USA    male          8
11  3-year-old     USA   total         16

如何旋转 pandas 数据框（有两个 headers）

How to pivot pandas dataframe (with two headers)

python

pivot

dataframe

pandas