我正在两个unix时间戳之间执行LAG操作，它工作得很好。但是，我想将结果列转换为小时/分钟
df_new = df_new.withColumn("diff", F.when(F.isnull(df_new.calculated_time - df_new.prev_value), 0)
.otherwise(df_new.calculated_time.cast('long') - df_new.prev_value.cast('long')))
输出-
+--------------------+--------------------+---------------+-------------+---------+
| primary_key| status|calculated_time| prev_value| diff|
+--------------------+--------------------+---------------+-------------+---------+
|{"approval_id": "...|Pending review by...| 1562315397258| null| 0|
|{"approval_id": "...| Denied| 1562936139570|1562315397258|620742312|
|{"approval_id": "...|Request clarifica...| 1563172343614|1562936139570|236204044|
|{"approval_id": "...| null| 1563172473488|1563172343614| 129874|
|{"approval_id": "...| Approved| 1563190166533|1563172473488| 17693045|
+--------------------+--------------------+---------------+-------------+---------+
在“差异”列中，我想有分钟。谁能帮忙吗？