pyspark中的 UDF

target_bundle_list = [["PP", 1,4],["PP", 17,4],["PP3", 18,4],["PP2", 1,4],["PP", 10,4]]

def getBundlesTarget(target_bundle_list):
    def temp3(country):
        target_bundle_list_ = [(i[1]) for i in target_bundle_list if i[0].upper() == "PP"]
        return [element for element in target_bundle_list_]
    return udf(temp3, ArrayType(IntegerType()))

增加的列为一个Array类型