Dataframe中按id将同一id的不同行变换成一行

问题：将如下图1按照s_info_windcode转换成图2

 

 

 

 

 

 

 解决方法：先对df进行s_info_windcode的groupby,然后进行unstack操作，columns重命名

方法一：代码简练，但有点难懂

 

 

分开展示

 

 

 

 

 

 方法二：容易理解，但偏机械

b = aa.copy()
df1 = pd.DataFrame(b.groupby('s_info_windcode')['f_prt_stocktotot'].apply(list))
stock = df1['f_prt_stocktotot'].apply(pd.Series).rename(columns=lambda x:'stocktotot'+str(x+1))  #axis=1 #apply(pd.Series)就会按列分开，每列重命名
stock.head()

df2 = pd.DataFrame(b.groupby('s_info_windcode')['f_prt_bondtotot'].apply(list))
bond = df2['f_prt_bondtotot'].apply(pd.Series).rename(columns=lambda x:'bond'+str(x+1))
bond.head()
pd.merge(stock,bond,on='s_info_windcode')

　　发现了自己之前的写法觉得好low

# cmfassetportfolio_1 = cmfassetportfolio[(cmfassetportfolio['f_prt_enddate']=='2019-09-30')]
# cmfassetportfolio_2 = cmfassetportfolio[(cmfassetportfolio['f_prt_enddate']=='2019-06-30')]
# cmfassetportfolio_3 = cmfassetportfolio[(cmfassetportfolio['f_prt_enddate']=='2019-03-31')]
# cmfassetportfolio_4 = cmfassetportfolio[(cmfassetportfolio['f_prt_enddate']=='2018-12-31')]
# cmfassetportfolio_5 = cmfassetportfolio[(cmfassetportfolio['f_prt_enddate']=='2018-09-30')]

# dt12 = pd.merge(cmfassetportfolio_1,cmfassetportfolio_2,on='s_info_windcode',how='outer',suffixes=('_1','_2'))
# dt34 = pd.merge(cmfassetportfolio_3,cmfassetportfolio_4,on='s_info_windcode',how='outer',suffixes=('_3','_4'))
# dt1234 = pd.merge(dt12,dt34,on='s_info_windcode',how='outer')
# dt = pd.merge(dt1234,cmfassetportfolio_5,on='s_info_windcode',how='outer')