我有一个Pandasdf，看起来像下面这样：

+---------+---------+------------+--------+
| Cluster | Country | Publishers | Assets |
+---------+---------+------------+--------+
| South   | IT      | SS         | Asset1 |
| South   | IT      | SS         | Asset2 |
| South   | IT      | SS         | Asset3 |
| South   | IT      | ML         | Asset1 |
| South   | IT      | ML         | Asset2 |
| South   | IT      | ML         | Asset3 |
| South   | IT      | TT         | Asset1 |
| South   | IT      | TT         | Asset2 |
| South   | IT      | TT         | Asset3 |
| South   | ES      | SS         | Asset1 |
| South   | ES      | SS         | Asset2 |
+---------+---------+------------+--------+

我想创建一个新列“Package”，该列使用基于以下列的累计计数：

• 出版商
• 资产

结果会是这样的：

+---------+---------+------------+--------+---------+
| Cluster | Country | Publishers | Assets | Package |
+---------+---------+------------+--------+---------+
| South   | IT      | SS         | Asset1 | 1       |
| South   | IT      | SS         | Asset2 | 1a      |
| South   | IT      | SS         | Asset3 | 1b      |
| South   | IT      | ML         | Asset1 | 2       |
| South   | IT      | ML         | Asset2 | 2a      |
| South   | IT      | ML         | Asset3 | 2b      |
| South   | IT      | TT         | Asset1 | 3       |
| South   | IT      | TT         | Asset2 | 3a      |
| South   | IT      | TT         | Asset3 | 3b      |
| South   | ES      | SS         | Asset1 | 4       |
| South   | ES      | SS         | Asset2 | 4a      |
+---------+---------+------------+--------+---------+

到目前为止我试过
df['Package'] = df.groupby（['Cluster'，'Publishers']）.cumcount（），但它似乎不起作用，因为在每个发布者示例完成后，值重置为0。