用特定于分组的值填充缺失值

用平均值去填充nan

s=pd.Series(np.random.randn(6))
s[::2]=np.nan
s

0         NaN
1   -0.118174
2         NaN
3   -0.308161
4         NaN
5   -0.655565
dtype: float64

s.fillna(s.mean())

stats=['Ohio','New York','Vermont','Florida','Oregon','Nevada','California','Idaho']
group_key=['East']*4 + ['West']*4
data=pd.Series(np.random.randn(8),index=stats)
data[['Vermont','Nevada','Idaho']]=np.nan
data

Ohio          0.345378
New York      1.538009
Vermont            NaN
Florida       0.113972
Oregon        1.618781
Nevada             NaN
California    0.728396
Idaho              NaN
dtype: float64

其中group_key是

group_key

['East', 'East', 'East', 'East', 'West', 'West', 'West', 'West']

data.groupby(group_key).mean()

East    0.665786
West    1.173589
dtype: float64

fill_mean=lambda g:g.fillna(g.mean())
data.groupby(group_key).apply(fill_mean)

Ohio          0.345378
New York      1.538009
Vermont       0.665786
Florida       0.113972
Oregon        1.618781
Nevada        1.173589
California    0.728396
Idaho         1.173589
dtype: float64

也可自定义字典作为填充值

fill_values={'East':0.5,'West':-1}
fill_func=lambda g:g.fillna(fill_values[g.name])
data.groupby(group_key).apply(fill_func)

Ohio          0.345378
New York      1.538009
Vermont       0.500000
Florida       0.113972
Oregon        1.618781
Nevada       -1.000000
California    0.728396
Idaho        -1.000000
dtype: float64

card_val=(list(range(1,11)) + [10]*3 )*4
card_val
#需要强制将rang转成list否则会报错

base_names=['A']+list(range(2,11))+['J','K','Q']
base_names

['A', 2, 3, 4, 5, 6, 7, 8, 9, 10, 'J', 'K', 'Q']

未完待续