Question

我想按时间顺序重新填充df中的行。

原始df：

     A   B   C Year
 0   ABC 0   A 1950
 1   CDE 1   A 1950
 2   XYZ 1   B 1954
 3   123 1   C 1954
 4   X12 1   B 1956
 5   123 1   D 1956
 6   124 1   D 1956

所需的df：

     A   B   C Year
 0   ABC 0   A 1950
 1   CDE 1   A 1950
 2   ABC 0   A 1951
 3   CDE 1   A 1951
 4   ABC 0   A 1952
 5   CDE 1   A 1952
 6   ABC 0   A 1953
 7   CDE 1   A 1953
 8   XYZ 1   B 1954
 9   123 1   C 1954
10   XYZ 1   B 1955
11   123 1   C 1955
12   X12 1   B 1956
13   123 1   D 1956
14   124 1   D 1956

我尝试将Year列转换为datetime，并使用逐年重采样进行向前填充。但这不起作用，因为如果按年份重新抽样，则每年仅重新抽样一次。

df.resample('YS').first().ffill().reset_index()

所需的df：

     A   B   C Year
 0   ABC 0   A 1950
 1   CDE 1   A 1950
 2   ABC 0   A 1951
 3   CDE 1   A 1951
 4   ABC 0   A 1952
 5   CDE 1   A 1952
 6   ABC 0   A 1953
 7   CDE 1   A 1953
 8   XYZ 1   B 1954
 9   123 1   C 1954
10   XYZ 1   B 1955
11   123 1   C 1955
12   X12 1   B 1956
13   123 1   D 1956
14   124 1   D 1956

Answer 1

我觉得这是一个unnesting问题

s=df.astype(str).groupby('Year').agg(list)
s.index=s.index.astype(int)
s1=s.reindex(np.arange(s.index.min(),s.index.max()+1),method='ffill')
yourdf=unnesting(s1,list('ABC')).reset_index()
yourdf
Out[117]: 
    Year    A  B  C
0   1950  ABC  0  A
1   1950  CDE  1  A
2   1951  ABC  0  A
3   1951  CDE  1  A
4   1952  ABC  0  A
5   1952  CDE  1  A
6   1953  ABC  0  A
7   1953  CDE  1  A
8   1954  XYZ  1  B
9   1954  123  1  C
10  1955  XYZ  1  B
11  1955  123  1  C
12  1956  X12  1  B
13  1956  123  1  D
14  1956  124  1  D

def unnesting(df, explode):
    idx = df.index.repeat(df[explode[0]].str.len())
    df1 = pd.concat([
        pd.DataFrame({x: np.concatenate(df[x].values)}) for x in explode], axis=1)
    df1.index = idx
    return df1.join(df.drop(explode, 1), how='left')

Answer 2

您可以尝试以下方法：

df_out = df.set_index([pd.to_datetime(df['Year'], format='%Y'),'A','B','C'])\
           .unstack([1,2,3]).resample('A').ffill()\
           .stack([1,2,3]).reset_index([1,2,3])

df_out = df_out.assign(Year=pd.to_datetime(df_out.index).year).reset_index(drop=True)
df_out

输出：

      A  B  C  Year
0   ABC  0  A  1950
1   CDE  1  A  1950
2   ABC  0  A  1951
3   CDE  1  A  1951
4   ABC  0  A  1952
5   CDE  1  A  1952
6   ABC  0  A  1953
7   CDE  1  A  1953
8   123  1  C  1954
9   XYZ  1  B  1954
10  123  1  C  1955
11  XYZ  1  B  1955
12  123  1  D  1956
13  X12  1  B  1956

Answer 3

我通过旋转和融化采取了不同的方法。似乎在工作。任何人都看到了问题。？

data = {'year': ['2000', '2000', '2005', '2005', '2007', '2007', '2007', '2009'],
'country':['UK', 'US', 'FR','US','UK','FR','US','UK'],
'sales': [10, 21, 20, 10,12,20, 10,12],
'rep': ['john', 'john', 'claire','claire', 'kyle','kyle','kyle','amy']
}
df=pd.DataFrame(data)


    year    country sales   rep
0   2000    UK  10  john
1   2000    US  21  john
2   2005    FR  20  claire
3   2005    US  10  claire
4   2007    UK  12  kyle
5   2007    FR  20  kyle
6   2007    US  10  kyle
7   2009    UK  12  amy

首先做一个枢轴...

dfp=pd.pivot_table(df,index=['country','rep'],values=['sales'],columns=['year']).fillna(0)
dfp=dfp.xs('sales', axis=1, drop_level=True)

    year    2000    2005    2007    2009
country rep             
FR  claire  0.0 20.0    0.0 0.0
kyle    0.0 0.0 20.0    0.0
UK  amy 0.0 0.0 0.0 12.0
john    10.0    0.0 0.0 0.0
kyle    0.0 0.0 12.0    0.0
US  claire  0.0 10.0    0.0 0.0
john    21.0    0.0 0.0 0.0
kyle    0.0 0.0 10.0    0.0

然后是一些逻辑，以复制列。

cols=dfp.columns.astype(int).values
dft=dfp.copy()
i=0
for col in cols :
    if col != cols[-1]:
        for newcol in range(col+1,cols[i+1]):
            dft[str(newcol)]=dft[str(col)]
    i+=1

    year    2000    2005    2007    2009    2001    2002    2003    2004    2006    2008
country rep                                     
FR  claire  0.0 20.0    0.0 0.0 0.0 0.0 0.0 0.0 20.0    0.0
kyle    0.0 0.0 20.0    0.0 0.0 0.0 0.0 0.0 0.0 20.0
UK  amy 0.0 0.0 0.0 12.0    0.0 0.0 0.0 0.0 0.0 0.0
john    10.0    0.0 0.0 0.0 10.0    10.0    10.0    10.0    0.0 0.0
kyle    0.0 0.0 12.0    0.0 0.0 0.0 0.0 0.0 0.0 12.0
US  claire  0.0 10.0    0.0 0.0 0.0 0.0 0.0 0.0 10.0    0.0
john    21.0    0.0 0.0 0.0 21.0    21.0    21.0    21.0    0.0 0.0
kyle    0.0 0.0 10.0    0.0 0.0 0.0 0.0 0.0 0.0 10.0

然后融化了它们，使其恢复为原始格式。

dfm=dft.reset_index()
dfm=dfm.melt(id_vars=['country','rep'],value_vars=dfm.columns.values[2:],var_name='Year',value_name='sales')
dfm=dfm.loc[dfm.sales>0].reset_index(drop='True')

    country rep Year    sales
0   UK  john    2000    10.0
1   US  john    2000    21.0
2   FR  claire  2005    20.0
3   US  claire  2005    10.0
4   FR  kyle    2007    20.0
5   UK  kyle    2007    12.0
6   US  kyle    2007    10.0
7   UK  amy     2009    12.0
8   UK  john    2001    10.0
9   US  john    2001    21.0
10  UK  john    2002    10.0
11  US  john    2002    21.0
12  UK  john    2003    10.0
13  US  john    2003    21.0
14  UK  john    2004    10.0
15  US  john    2004    21.0
16  FR  claire  2006    20.0
17  US  claire  2006    10.0
18  FR  kyle    2008    20.0
19  UK  kyle    2008    12.0
20  US  kyle    2008    10.0

重复pandas数据框中的行，列值顺序更改

3 个答案: