我有下面的数据框，经过几次计算后得到的结果。
+------+--------+------------+------------+------------+-----------+-----------------+------------+-----------+-----------------+
| label| machine|value1_count|value2_count|value1_score|value1_band|value1_band_score|value2_score|value2_band|value2_band_score|
+------+--------+------------+------------+------------+-----------+-----------------+------------+-----------+-----------------+
|label1|machine1| 2| 0| 91.67| low| 70| 100.0| normal| 100|
|label1|machine2| 1| 1| 95.83| low| 70| 95.83| low| 70|
|label2|machine3| 3| 2| 87.5| low| 70| 91.67| low| 70|
|label2|machine4| 1| 1| 95.83| low| 70| 95.83| low| 70|
+------+--------+------------+------------+------------+-----------+-----------------+------------+-----------+-----------------+
数据帧可以具有value3，value4等的_count，_score，_band，_band_score列。现在，我想取消透视此数据，以便获得以下格式的数据：
+------+--------+----------+-----+-----+------+----------+
| label| machine|value_type|count|score| band|band_score|
+------+--------+----------+-----+-----+------+----------+
|label1|machine1| value1| 2|91.67| low| 70|
|label1|machine1| value2| 0| 100|normal| 100|
|label1|machine2| value1| 1|95.83| low| 70|
|label1|machine2| value2| 1|95.83| low| 70|
|label2|machine3| value1| 3| 87.5| low| 70|
|label2|machine3| value2| 2|91.67| low| 70|
|label2|machine4| value1| 1|95.83| low| 70|
|label2|machine4| value2| 1|95.83| low| 70|
+------+--------+----------+-----+-----+------+----------+
我认为在这里不使用正则表达式会很有用，但我无法实现。有人可以帮我吗？
预先感谢