基於 Python Z251D2BBFE9A3B78EAZ5DC69

Question

我有一個像這樣的 DataFrame：

 category    uid sales_1 sales_2
0    Grocery     1   XX   XX
1    Grocery     2   XX   ZZ
2    Sports      3   XX   ZZ
3    Grocery     4   ZZ   XX
4    Beauty      5   ZZ   ZZ
5    Beauty      6   ZZ   ZZ
6    Sports      7   ZZ   XX
7    Grocery     8   ZZ   XX
...

我需要將 sales_1 列與 sales_2 列進行比較。 比較結果將反映在第一和第二個新列中。 如果 sales_1 == sales_2 則這 2 個新列中的值應為“無更改”和“確定”。 如果 sales_1.= sales_2 的值應該是“改變”和“差距”：最后我想要一個以下數據框：

 category    uid sales_1 sales_2  first     second
0    Grocery     1   XX   XX    no changes  OK
1    Grocery     2   XX   ZZ    changed     gap
2    Sports      3   XX   ZZ    changed     gap
3    Grocery     4   ZZ   XX    changed     gap
4    Beauty      5   ZZ   ZZ    no changes  OK
5    Beauty      6   ZZ   ZZ    no changes  OK
6    Sports      7   ZZ   XX    changed     gap
7    Grocery     8   ZZ   XX    changed     gap
...

我真的很感激任何建議。

Answer 1

您可以使用 numpy 中的where() function ：

df['first'] = np.where(df.sales_1 == df.sales_2, 'no changes', 'changed')
df['second'] = np.where(df.sales_1 == df.sales_2, 'OK', 'gap')

Answer 2

您可以首先為first列和second列分配一個默認值，然后根據銷售是否發生變化來應用過濾。

import pandas as pd

df = pd.DataFrame(
    {
        'category': ['Grocery', 'Sports', 'Beauty'],
        'sales_1': ['XX', 'ZZ', 'XX'],
        'sales_2': ['XX', 'XY', 'ZZ'],
    }
)

changed_sales = df['sales_1'] != df['sales_2']

df['first'] = 'no changes'
df.loc[changed_sales, 'first'] = 'changed'
df['second'] = 'OK'
df.loc[changed_sales, 'second'] = 'gap'

print(df)

Output

  category sales_1 sales_2       first second
0  Grocery      XX      XX  no changes     OK
1   Sports      ZZ      XY     changed    gap
2   Beauty      XX      ZZ     changed    gap

Answer 3

你可以使用列表理解

df['first']= ["no changes" if s1 == s2 else "changed" for (s1, s2) in zip(df['sales_1'], df['sales_2']) ]
df['second'] = ["OK" if s1 == s2 else "gap" for (s1, s2) in zip(df['sales_1'], df['sales_2']) ]