Python 在交叉表操作中，同时设置colnames=[None]和margins=True会导致错误

我有一个由两个分类变量组成的简单数据框架。这可以简化如下：

tempDF = pd.DataFrame({ 'species': ["cat","dog","dog","dog","cat","cat","cat","dog","cat","cat"],
                        'gender':  ["female","female","female","male","male","female","female","female","male","female"]})

print(tempDF)

       gender species
    0  female     cat
    1  female     dog
    2  female     dog
    3    male     dog
    4    male     cat
    5  female     cat
    6  female     cat
    7  female     dog
    8    male     cat
    9  female     cat

tempTable = pd.crosstab(tempDF['species'],tempDF['gender'])

gender   female  male
species              
cat           4     2
dog           3     1

我可以创建一个简单的交叉表格，如下所示：

tempDF = pd.DataFrame({ 'species': ["cat","dog","dog","dog","cat","cat","cat","dog","cat","cat"],
                        'gender':  ["female","female","female","male","male","female","female","female","male","female"]})

print(tempDF)

       gender species
    0  female     cat
    1  female     dog
    2  female     dog
    3    male     dog
    4    male     cat
    5  female     cat
    6  female     cat
    7  female     dog
    8    male     cat
    9  female     cat

tempTable = pd.crosstab(tempDF['species'],tempDF['gender'])

gender   female  male
species              
cat           4     2
dog           3     1

但是，我希望种类是列中的变量，而不是索引，以便可以重置索引：

tempTable = tempTable.reset_index(drop=False)

gender species  female  male
0          cat       4     2
1          dog       3     1

但现在看来，该索引被标记为“性别”，这令人困惑，因为“男性”和“女性”已经是每一列的标题。事实上，“性别”一词实际上似乎是列标题的名称，可以设置为任何内容，包括无：

tempTable.columns.name = None

  species  female  male
0     cat       4     2
1     dog       3     1

这正是我想要的

事实上，这可以在初始交叉表阶段更有效地完成：

tempTable = pd.crosstab(tempDF['species'],tempDF['gender'],colnames=[None])

         female  male
species              
cat           4     2
dog           3     1

（注意，似乎不需要通过方括号中的列表来产生任何影响。但无论如何。）

如果我还希望在交叉表格中包含边距，则会出现问题：

tempTable = pd.crosstab(tempDF['species'],tempDF['gender'],colnames=[None],margins=True)

这一次，我得到：

KeyError: 'Level None not found'

包括colnames=[None]或margins=True都可以。但是为什么include margins=True应该与colnames=[None]交互以产生错误呢？有没有办法将列名设置为“无”并在同一命令中包含边距？还是我必须完成上面列出的所有步骤