我创建了一个泛型函数，根据所需的数字删除前导“0”：

from pyspark.sql import functions as F
def remove_lead_zero(col, n):
    """
    col: name of the column you want to modify
    n: number of leading 0 you want to remove
    """
    return F.when(
        F.regexp_extract(col, "^0{{{n}}}[^0]".format(n=n), 0) != "",
        F.expr("substring({col}, {n}, length({col}))".format(col=col, n=n+1))
    ).otherwise(F.col(col))
df.withColumn("output", remove_lead_zero("subcategory", 2)).show()
+-----------+-------+
|subcategory| output|
+-----------+-------+
|      00EEE|    EEE|
|    0000EEE|0000EEE|
|       0EEE|   0EEE|
+-----------+-------+
df.withColumn("output", remove_lead_zero("subcategory", 1)).show()
+-----------+-------+
|subcategory| output|
+-----------+-------+
|      00EEE|  00EEE|
|    0000EEE|0000EEE|
|       0EEE|    EEE|
+-----------+-------+