主页 / coding / 问题 / 78932699
Accepted
elokema
Asked: 2024-08-30 23:08:48 +0800 CST

Pyspark：编写一个通用函数

  • 772

我想在函数pyspark中写入这部分

df = (df.withColumn("January", F.lit(None).cast('double'))
        .withColumn("February", F.lit(None).cast('double'))
        .withColumn("March", F.lit(None).cast('double'))
        .withColumn("April", F.lit(None).cast('double'))
        .withColumn("May", F.lit(None).cast('double'))
        .withColumn("June", F.lit(None).cast('double'))
        .withColumn("July", F.lit(None).cast('double'))
        .withColumn("August", F.lit(None).cast('double'))
        .withColumn("September", F.lit(None).cast('double'))
        .withColumn("November", F.lit(None).cast('double'))
        .withColumn("December", F.lit(None).cast('double'))
pyspark

1 个回答

  1. Best Answer

    您可以使用withColumns代替组withColumn

    months = ["January", ... , "December"]
df = df.withColumns(
    {month: F.lit(None).cast('double') for month in months}
)

    文档在这里： https：//spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.DataFrame.withColumns.html

    • 1

相关问题