Joining two DataFrames in Spark SQL and selecting columns of only one

Question

I have two DataFrames in Spark SQL (D1 and D2).

I am trying to inner join both of them D1.join(D2, "some column") and get back data of only D1, not the complete data set.

Both D1 and D2 are having the same columns.

Could some one please help me on this?

I am using Spark 1.6.

Answer 1

Let say you want to join on "id" column. Then you could write :

val sqlContext = new org.apache.spark.sql.SQLContext(sc)
import sqlContext.implicits._    
d1.as("d1").join(d2.as("d2"), $"d1.id" === $"d2.id").select($"d1.*")

Answer 2

As an alternate answer, you could also do the following without adding aliases:

d1.join(d2, d1("id") === d2("id"))
  .select(d1.columns.map(c => d1(c)): _*)

Answer 3

You could use left_semi:

d1.as("d1").join(d2.as("d2"), $"d1.id" === $"d2.id", "left_semi")

Semi-join takes only rows from the left dataset where joining condition is met.

There's also another interesting join type: left_anti, which works similarily to left_semi but takes only those rows where the condition is not met.

Answer 4

You should use leftsemi join which is similar to inner join difference being leftsemi join returns all columns from the left dataset and ignores all columns from the right dataset.

You can try something like the below in Scala to Join Spark DataFrame using leftsemi join types.

empDF.join(deptDF,empDF("emp_dept_id") ===  deptDF("dept_id"),"leftsemi")
    .show(false)

If you are using Python use below PySpark join dataframe example

empDF.join(deptDF,empDF.emp_dept_id ==  deptDF.dept_id,"leftsemi") \
   .show(truncate=False)