过滤pyspark dataframe以获取最近N天的行不能正常工作，并返回一些早于指定日期的数据

from pyspark.sql import SparkSession
from pyspark.sql.functions import col, to_date
from datetime import datetime, timedelta

# 创建SparkSession
spark = SparkSession.builder.getOrCreate()

# 创建示例数据
data = [("2022-01-01", "data1"),
        ("2022-01-02", "data2"),
        ("2022-01-03", "data3"),
        ("2022-01-04", "data4"),
        ("2022-01-05", "data5")]

df = spark.createDataFrame(data, ["date", "data"])

# 将日期列转换为日期类型
df = df.withColumn("date", to_date(col("date")))

# 获取当前日期
current_date = datetime.now().date()

# 计算N天前的日期
n_days_ago = current_date - timedelta(days=N)

# 过滤数据
filtered_df = df.filter(col("date") >= n_days_ago)

# 返回早于指定日期的数据
earlier_than_date_df = df.filter(col("date") < specified_date)

# 显示结果
filtered_df.show()
earlier_than_date_df.show()