如何循环遍历dataFrame的每一行，并根据条件删除该行

Question

I'm beginner in Spark & scala. I would like to know on how to loop through each row of dataFrame, and remove the row based on a condition.

Answer 1

You can use filter operation on a dataframe in which you can specify a condition on the basis of which you want to filter record. Below is an example:

import org.apache.spark.sql.SparkSession
import org.apache.spark.sql.{DataFrame, functions => F}


object Example extends App {

  val spark = SparkSession.builder.appName("Simple Application").master("local")
    .getOrCreate()

  import spark.implicits._

  val df1 = spark.sparkContext.parallelize(
    List(
      ("Cust1", "Prod1", "Promo1", 1),
      ("Cust1", "Prod1", "Promo2", 2),
      ("Cust2", "Prod5", "Promo4", 11),
      ("Cust2", "Prod8", "Promo4", 12),
      ("Cust3", "Prod3", "Promo9", 14),
      ("Cust3", "Prod2", "Promo6", 13)
    )).toDF("customer", "product", "promotion", "cardid")

   .show()

}

The output of above code is:

+--------+-------+---------+------+
|customer|product|promotion|cardid|
+--------+-------+---------+------+
|   Cust1|  Prod1|   Promo1|     1|
|   Cust1|  Prod1|   Promo2|     2|
+--------+-------+---------+------+

In the above example i filtered records where value in product column is "Prod1" as can be seen in: df1.filter(F.col("product") === "Prod1")

Filter operation iterates on each row of the dataframe and check the condition provided and keep all records where the condition is true.