PySpark将字符串列转换为日期时间类型

from pyspark.sql import SparkSession
from pyspark.sql.functions import to_date, to_timestamp

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

# 创建示例数据集
data = [("2022-01-01", "2022-01-01 12:00:00"),
        ("2022-02-01", "2022-02-01 12:00:00"),
        ("2022-03-01", "2022-03-01 12:00:00")]

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

# 将字符串列转换为日期类型
df = df.withColumn("date", to_date(df.date_str))

# 将字符串列转换为时间戳类型
df = df.withColumn("timestamp", to_timestamp(df.timestamp_str))

# 显示结果
df.show()

+----------+-------------------+----------+-------------------+
| date_str |   timestamp_str   |   date   |     timestamp     |
+----------+-------------------+----------+-------------------+
|2022-01-01|2022-01-01 12:00:00|2022-01-01|2022-01-01 12:00:00|
|2022-02-01|2022-02-01 12:00:00|2022-02-01|2022-02-01 12:00:00|
|2022-03-01|2022-03-01 12:00:00|2022-03-01|2022-03-01 12:00:00|
+----------+-------------------+----------+-------------------+