根据 Python 中的条件更改现有列中的值

Question

嗨，对于每个分组（基于动物和颜色），我想在检查列中检查 Action = 'See' 是否为 'No'。如果 Action = 'See' 为 'No'，则 Check 列中的特定分组将是 'No'。

如果 Action = See and Check = Yes，则保持旧值（参考 Tiger, White）

animallist= df['Animal'].unique().tolist()
colourlist= df['Colour'].unique().tolist()

for i in animallist:
   for j in colourlist:
       df = df[(df['Animal'] == i) & (df['Colour'] == j)]
       df['Check'] = np.where((df['Action'] == 'See') & (df['Check'] == 'no'), 'no', 
                     df['Check'])

Answer 1

我认为您想要的是：更改“正确？”的值仅当某些其他列具有特定值时才列。否则，请留下“正确？”单独列。您可以创建各种测试来更改列值。这里有两个例子：

# if Action == View and Colour = Red, then set correct to yes
# regardless of what Correct? was before

df.loc[(df['Action'] == 'View') & (df['Colour'] == 'Red'), 'Correct?'] = 'Yes'

# similar to above, but change Correct? only if it was not already 'Yes'

df.loc[(df['Action'] == 'View') & (df['Colour'] == 'Red')  & (df['Correct?'] == 'No'), 'Correct?'] = 'Yes'

# create a new correction column, default to No
# change to yes only under certain conditions

df['new_correct'] = 'No'
df.loc[(df['Action'] == 'View') & (df['Colour'] == 'Red'), 'new_correct'] = 'Yes'

要查找具有特定标准的所有动物，然后根据该标准更改主 df，您可以执行以下操作：

df_no = df[(df['Colour'] == 'Red') & (df['Action'] == 'View') & (df['Correct?'] == 'No')].drop_duplicates()

df_no = df_no[[df_no['Animal']]   # keep only the 'Animal' column

df = df.merge(df_no, on='Animal', how='left', indicator = 'True').
df.loc[df['_merge'] == 'both', 'Correct?'] = 'no'

你可以用同样的想法做更复杂的事情。例如，您需要弄清楚一个动物是否同时具有“是”和“否”来表示“正确”？那你想做什么。