您需要使用or条件

test_df = df[(df['purchase_qty'] != 0) | (df['other_adjustments'] != 0) | (df['sold'] != 0)]

或者你可以做一个逆运算

test_df = df[~((df['purchase_qty'] == 0) & (df['other_adjustments'] == 0) & (df['sold'] == 0))]

如果至少有一个值不是0，则可以过滤3列-因此删除3列中所有0的行：

test_df = df[df[['purchase_qty','other_adjustments','sold']].ne(0).any(axis=1)] 
print (test_df)
    item purchase_date  purchase_qty  purchase_price  other_adjustments  \
1  50064       2022-01             0             0.0                 -5   
2  50066       2022-01             0             0.0               2979   
3  50202       2022-01             0             0.0              14673   
4  50204       2022-01             0             0.0               2538   

     sold  
1   854.0  
2     0.0  
3  1320.0  
4     0.0

你可以试试这个：
test_df = df[(df['purchase_qty'] != 0) | (df['other_adjustments'] != 0) | (df['sold'] != 0)]
此状态：如果这些值中有一个不为0，则保留它，否则过滤掉。

这个也行！！

df = df[(df[['purchase_qty', 'other_adjustments', 'sold']] != 0).any(axis=1)]

这是对所用方法的修复：
df = df.drop(df[(df.purchase_qty == 0) & (df.other_adjustments == 0) & (df.sold == 0.0)].index)