groupby 分组统计
1.根据某些条件将数据分组
2.对每个组独立应用函数
3.将结果合并到一个数据结构中
Dataframe在行或列上分组，将一个函数应用到各个分组并产生一个新值，然后函数执行结果被合并到最终的结果对象中
#分组 import numpy as npimport pandas as pddf = pd.DataFrame({'A':['foo','bar','foo','bar','foo','bar','foo','foo'], 'B':['one','one','two','three','two','two','one','three'], 'C':np.random.randn(8), 'D':np.random.randn(8)}) print(df) print('------') print(df.groupby('A'),type(df.groupby('A'))) #直接分组得到的是groupby对象，是一个中间数据，没有进行计算 print(df.groupby('A').sum())#自动过滤字符串列 print(df.groupby('A').mean())#平均值 b = df.groupby(['A','B']).mean() print(b,type(b),'\n',b.columns) c = df.groupby(['A'])['D'].mean()#以A分组，取D列平均值 print(c,type(c),'\n')结果: A B C D0 foo one 0.429615 -0.7087821 bar one 0.891751 1.1405752 foo two -0.261858 -0.5168353 bar three 1.310361 0.2696574 foo two 1.048076 1.3742185 bar two -0.410148 1.0611326 foo one -1.124137 -0.7293677 foo three 0.289513 0.892714------<pandas.core.groupby.DataFrameGroupBy object at 0x000000000FBACA58> <class 'pandas.core.groupby.DataFrameGroupBy'> C DA bar 1.791963 2.471364foo 0.381208 0.311947 C DA bar 0.597321 0.823788foo 0.076242 0.062389 C DA B bar one 0.891751 1.140575 three 1.310361 0.269657 two -0.410148 1.061132foo one -0.347261 -0.719074 three 0.289513 0.892714 two 0.393109 0.428691 <class 'pandas.core.frame.DataFrame'> Index(['C', 'D'], dtype='object')Abar 0.823788foo 0.062389Name: D, dtype: float64 <class 'pandas.core.series.Series'>
#分组 - 可迭代的对象 df = pd.DataFrame({'X':['A','B','A','B'],'Y':[1,3,4,2]}) print(df) print(df.groupby('X'),type(df.groupby('X'))) print('-------') print(list(df.groupby('X')),'->可迭代对象，直接生成list\n') print(list(df.groupby('X'))[0],'->以元组的形式显示') for n,g in df.groupby('X'): print(n) print(g) print('###') print('--------') #n是组名，g是分组后的DataFrame print(df.groupby(['X']).get_group('A'),'\n') print(df.groupby(['X']).get_group('B'),'\n') #.get_group提取分组后的组 grouped = df.groupby(['X']) print(grouped.groups) print(grouped.groups['A'])#也可写 df.groupby('X').groups['A'] print('-------') #.groups：将分组后的groups转化为dict #可以字典索引方法来查看groups里的元素 sz = grouped.size() print(sz,type(sz)) #.size() 查看分组后的长度 print('---------') df = pd.DataFrame({'A':['foo','bar','foo','bar','foo','bar','foo','foo'], 'B':['one','one','two','three','two','two','one','three'], 'C':np.random.randn(8), 'D':np.random.randn(8)}) grouped = df.groupby(['A','B']).groups print(df) print(grouped) print(grouped['foo','three']) dic=dict({'A':[1,2,3], 'B':[2,3,4]}) print(dic,type(dic))结果： X Y0 A 11 B 32 A 43 B 2<pandas.core.groupby.DataFrameGroupBy object at 0x000000000F889F60> <class 'pandas.core.groupby.DataFrameGroupBy'>-------[('A', X Y0 A 12 A 4), ('B', X Y1 B 33 B 2)] ->可迭代对象，直接生成list('A', X Y0 A 12 A 4) ->以元组的形式显示A X Y0 A 12 A 4###B X Y1 B 33 B 2###-------- X Y0 A 12 A 4 X Y1 B 33 B 2 {'A': Int64Index([0, 2], dtype='int64'), 'B': Int64Index([1, 3], dtype='int64')}Int64Index([0, 2], dtype='int64')-------XA 2B 2dtype: int64 <class 'pandas.core.series.Series'>--------- A B C D0 foo one -0.881923 -0.8251021 bar one -0.626412 -0.6186382 foo two -1.741248 1.5576983 bar three 1.076928 1.7382654 foo two -0.954103 -0.7414155 bar two 1.224841 -0.4794726 foo one 0.680046 -0.4761377 foo three -1.519952 -0.421738{('bar', 'one'): Int64Index([1], dtype='int64'), ('bar', 'three'): Int64Index([3], dtype='int64'), ('bar', 'two'): Int64Index([5], dtype='int64'), ('foo', 'one'): Int64Index([0, 6], dtype='int64'), ('foo', 'three'): Int64Index([7], dtype='int64'), ('foo', 'two'): Int64Index([2, 4], dtype='int64')}Int64Index([7], dtype='int64'){'A': [1, 2, 3], 'B': [2, 3, 4]} <class 'dict'>
#其他轴上分组 df = pd.DataFrame({'data1':np.random.randn(2), 'data2':np.random.randn(2), 'key1':['a','b'], 'key2':['one','two']}) print(df) print(df.dtypes) print('--------') for n,p in df.groupby(df.dtypes,axis=1): print(n) print(p) print('##') #按照值类型分组，分为2组结果: data1 data2 key1 key20 0.813374 0.232957 a one1 -0.213256 1.393156 b twodata1 float64data2 float64key1 objectkey2 objectdtype: object--------float64 data1 data20 0.813374 0.2329571 -0.213256 1.393156##object key1 key20 a one1 b two##
#通过字典或者Series分组 df = pd.DataFrame(np.arange(16).reshape(4,4), columns = ['a','b','c','d']) print(df) print('-------') mapping = {'a':'one','b':'one','c':'two','d':'two','e':'three'} print(mapping) by_column = df.groupby(mapping,axis = 1) print(by_column.sum()) print('---------') #mapping中 a,b列对应为one，c,d列对应为two，以字典为分组 s=pd.Series(mapping) print(s) print(s.groupby(s).count()) #s中，index = a，b对应的是one；c，d对应的是two，以Series来分组结果: a b c d0 0 1 2 31 4 5 6 72 8 9 10 113 12 13 14 15-------{'a': 'one', 'b': 'one', 'c': 'two', 'd': 'two', 'e': 'three'} one two0 1 51 9 132 17 213 25 29---------a oneb onec twod twoe threedtype: objectone 2three 1two 2dtype: int64
