将现有列替换为多个列的相同条件

Question

下面是我的pandas数据帧，每列包含0或1。如果其他列（severe_toxic，…，identity hate）中至少有一列包含1，我正在尝试将现有的'toxic'替换为1

我尝试了下面的代码，但它给出了错误

我试过的代码：

# a1 - above dataframe's name
a1['toxic'] = [1 if any(a1[[severe_toxic','obscene','threat','insult','identity_hate']]) ==1]

Answer 1

使用Pandas而不是Python中的any：

cols = ['severe_toxic', 'obscene', 'threat', 'insult', 'identity_hate']
a1['toxic'] = a1[cols].any(axis=1).astype(int)

Answer 2

使用：

df['toxic'] = np.where((df[df.columns[1:]]==1).any(axis=1), 1, df['toxic'])

Input:

   toxic  severe_toxic  obscene  threat  insult  identity_hate
0      0             0        0       0       0              0
1      0             0        0       0       0              0
2      0             0        0       0       0              1
3      0             0        0       0       0              1
4      0             0        0       0       0              0

Output:

   toxic  severe_toxic  obscene  threat  insult  identity_hate
0      0             0        0       0       0              0
1      0             0        0       0       0              0
2      1             0        0       0       0              1
3      1             0        0       0       0              1
4      0             0        0       0       0              0

Setup:

df = pd.DataFrame(data={'toxic':[0]*5,
                        'severe_toxic':[0]*5,
                        'obscene':[0]*5,
                        'threat':[0]*5,
                        'insult':[0]*5,
                        'identity_hate':[0,0,1,1,0]})