Python 使用基准年计算指数

df

我如何计算每种水果的真实指数后的年度指数列？基准年由索引_值==100的行给出

我试过：

fruit  year price index_value Boolean  index
apple  1960  11    
apple  1961  12    100        True
apple  1962  13    
apple  1963  13    100        True
banana 1960  11    
banana 1961  12    

预期产出：

df['index'] = df.groupby('fruit')['price'].apply(lambda x: (x/x.iloc[0] * 100).round(0))

---

我冒昧地为apple 1964 11调整了一行输入数据，以匹配您的输出示例。列

Boolean

是冗余的

fruit  year price index_value Boolean  index
apple  1960  11    
apple  1961  12    100        True      100
apple  1962  13                         108
apple  1963  13    100        True      100
apple  1964  11                         84 
banana 1961  12    

输出：

要获得所需的输出，首先为给定索引值之后的值创建子组

    fruit  year  price  index_value
0   apple  1960     11          NaN
1   apple  1961     12        100.0
2   apple  1962     13          NaN
3   apple  1963     13        100.0
4   apple  1964     11          NaN
5  banana  1960     11          NaN
6  banana  1961     12          NaN

输出：

然后可以计算索引值的百分比变化

    fruit  year  price  index_value  groups
0   apple  1960     11          NaN       0
1   apple  1961     12        100.0       1
2   apple  1962     13          NaN       1
3   apple  1963     13        100.0       2
4   apple  1964     11          NaN       2
5  banana  1960     11          NaN       0
6  banana  1961     12          NaN       0

输出：

df['groups'] = df.index_value.notna().groupby(df.fruit).cumsum().astype('int')
print(df)

    fruit  year  price  index_value  groups
0   apple  1960     11          NaN       0
1   apple  1961     12        100.0       1
2   apple  1962     13          NaN       1
3   apple  1963     13        100.0       2
4   apple  1964     11          NaN       2
5  banana  1960     11          NaN       0
6  banana  1961     12          NaN       0

df['index_change'] = (
    df[df.groups.ne(0)]
      .groupby(['fruit','groups'])['price'].apply(lambda x: np.floor((x/x.iloc[0] * 100)))
)
print(df)

    fruit  year  price  index_value  groups  index_change
0   apple  1960     11          NaN       0           NaN
1   apple  1961     12        100.0       1         100.0
2   apple  1962     13          NaN       1         108.0
3   apple  1963     13        100.0       2         100.0
4   apple  1964     11          NaN       2          84.0
5  banana  1960     11          NaN       0           NaN
6  banana  1961     12          NaN       0           NaN