Pandas-高级处理（五）：一般性“拆分-应用-合并”【一般化Groupby方法：apply】

一、一般化Groupby方法：apply

import numpy as np
import pandas as pd

# 一般化Groupby方法：apply

df = pd.DataFrame({'data1': np.random.rand(5),
                   'data2': np.random.rand(5),
                   'key1': list('aabba'),
                   'key2': ['one', 'two', 'one', 'two', 'one']})

# apply直接运行其中的函数
# 这里为匿名函数，直接描述分组后的统计量
x1 = df.groupby('key1').apply(lambda x: x.describe())
print("x1 = df.groupby('key1').apply(lambda x: x.describe()) = \n", x1)
print('-' * 200)


# f_df1函数：返回排序后的前n行数据
def f_df1(d, n):
    return (d.sort_index()[:n])


# f_df2函数：返回分组后表的k1列，结果为Series，层次化索引
def f_df2(d, k1):
    return (d[k1])


print("df = \n", df)
print('-' * 200)

x1 = df.groupby('key1').apply(f_df1, 2)
print("x1 = df.groupby('key1').apply(f_df1, 2) = \n", x1)
print('-' * 200)

x2 = df.groupby('key1').apply(f_df2, 'data2')
print("x2 = df.groupby('key1').apply(f_df2, 'data2') = \n", x2)
print('-' * 50)
print("type(x2) = ", type(x2))
print('-' * 200)

打印结果：

x1 = df.groupby('key1').apply(lambda x: x.describe()) = 
                data1     data2
key1                          
a    count  3.000000  3.000000
     mean   0.776583  0.498758
     std    0.357791  0.257671
     min    0.363582  0.225764
     25%    0.668674  0.379274
     50%    0.973767  0.532783
     75%    0.983083  0.635255
     max    0.992400  0.737726
b    count  2.000000  2.000000
     mean   0.593805  0.531914
     std    0.041973  0.451099
     min    0.564126  0.212939
     25%    0.578965  0.372426
     50%    0.593805  0.531914
     75%    0.608644  0.691402
     max    0.623484  0.850889
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
df = 
       data1     data2 key1 key2
0  0.363582  0.225764    a  one
1  0.973767  0.737726    a  two
2  0.623484  0.212939    b  one
3  0.564126  0.850889    b  two
4  0.992400  0.532783    a  one
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
x1 = df.groupby('key1').apply(f_df1, 2) = 
            data1     data2 key1 key2
key1                                
a    0  0.363582  0.225764    a  one
     1  0.973767  0.737726    a  two
b    2  0.623484  0.212939    b  one
     3  0.564126  0.850889    b  two
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
x2 = df.groupby('key1').apply(f_df2, 'data2') = 
 key1   
a     0    0.225764
      1    0.737726
      4    0.532783
b     2    0.212939
      3    0.850889
Name: data2, dtype: float64
--------------------------------------------------
type(x2) =  <class 'pandas.core.series.Series'>
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Process finished with exit code 0