如何在pySpark中将带有尾随破折号的数字格式化为负数？

from pyspark.sql.functions import regexp_replace

data = [("123-",), ("456-",), ("789",)]
df = spark.createDataFrame(data, ["number"])

df = df.withColumn("formatted_number", regexp_replace("number", "-$", ""))

df.show()

+------+----------------+
|number|formatted_number|
+------+----------------+
|  123-|            -123|
|  456-|            -456|
|   789|             789|
+------+----------------+