ComeçarComece de graça

Aggregate dot SQL

The following code uses SQL to set the value of a dataframe called df.

df = spark.sql("""
SELECT *, 
LEAD(time,1) OVER(PARTITION BY train_id ORDER BY time) AS time_next 
FROM schedule
""")
  • The LEAD clause has an equivalent function in pyspark.sql.functions.
  • The PARTITION BY, and ORDER BY clauses each have an equivalent dot notation function that is called on the Window object.
  • The following imports are available:
    • from pyspark.sql import Window
    • from pyspark.sql.functions import lead

Este exercício faz parte do curso

Introduction to Spark SQL in Python

Ver curso

Instruções do exercício

  • Create a dataframe called dot_df that contains the identical result as df, using dot notation instead of SQL.

Exercício interativo prático

Experimente este exercício completando este código de exemplo.

# Obtain the identical result using dot notation 
dot_df = df.withColumn('time_next', ____('time', 1)
        .over(____.____('train_id')
        .____('time')))
Editar e executar o código