Search code examples
apache-sparkrowsusinggenerate

How do I Generate rows depending upon the column value in spark?

suppose If I have single column with one row 

+---+
| id|
+---+
|  4|
+---+

then how do i generate rows depending up on the value of a column

+---+
| id|
+---+
| 1 |
|---|
| 2 |
|---|
| 3 |
|---|
| 4 |
+---+
Solution

  • You can define a udf function for that to generate range and then use explode function to make them to separate rows 

    import org.apache.spark.sql.functions._
def generateUdf = udf((column: Int)=> (1 to column).toArray)

df.withColumn("id", explode(generateUdf(col("id")))).show(false)

    which should give you 

    +---+
|id |
+---+
|1  |
|2  |
|3  |
|4  |
+---+