'Merging the similar column names while joining two dataframes using pyspark
In the below program ,the duplicate columns are getting created while joining two dataframes in pyspark .
>>> spark = SparkSession.builder.appName("Join").getOrCreate()
>>> dict=[{"Emp_id" : 123 , "Emp_name" : "Raja" }, {"Emp_id" : 456 , "Emp_name" : "Ravi"}]
>>> dict1=[{"Emp_id" : 123 , "Dep_name" : "Computer" } , {"Emp_id" : 456 ,"Dep_name" :"Economy"}]
>>> df=spark.createDataFrame(dict)
>>> df1=spark.createDataFrame(dict1)
>>> df2=df.join(df1,df.Emp_id == df1.Emp_id, how = 'inner')
>>> df.show()
+------+--------+
|Emp_id|Emp_name|
+------+--------+
| 123| Raja|
| 456| Ravi|
+------+--------+
>>> df1.show()
+--------+------+
|Dep_name|Emp_id|
+--------+------+
|Computer| 123|
| Economy| 456|
+--------+------+
>>> df2=df.join(df1,df.Emp_id == df1.Emp_id, how = 'inner')
>>> df2.show()
+------+--------+--------+------+
|Emp_id|Emp_name|Dep_name|Emp_id|
+------+--------+--------+------+
| 123| Raja|Computer| 123|
| 456| Ravi| Economy| 456|
+------+--------+--------+------+
Is there any other way to get the data like below as the result of join with overwriting the columns as similar as in SAS?
+------+--------+--------+
|Emp_id|Emp_name|Dep_name|
+------+--------+--------+
| 123| Raja|Computer|
| 456| Ravi| Economy|
+------+--------+--------+
Solution 1:[1]
In your join condition replace df.Emp_id == df1.Emp_id
with ['Emp_id']
df2=df.join(df1,['Emp_id'], how = 'inner')
df2.show()
#+------+--------+--------+
#|Emp_id|Emp_name|Dep_name|
#+------+--------+--------+
#| 123| Raja|Computer|
#| 456| Ravi| Economy|
#+------+--------+--------+
Solution 2:[2]
when joinin two dataframes on same column, explicitly specify join column on which you want to apply join in 'on' clouse.
df2=df.join(df1, on='Emp_id' how = 'inner')
df2.show()
#+------+--------+--------+
#|Emp_id|Emp_name|Dep_name|
#+------+--------+--------+
#| 123| Raja|Computer|
#| 456| Ravi| Economy|
#+------+--------+--------+
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
Solution | Source |
---|---|
Solution 1 | notNull |
Solution 2 | MOHD NAYYAR |