Pyspark : Change nested column datatype

Question

How can we change the datatype of a nested column in Pyspark? For rxample, how can I change the data type of value from string to int?

Reference:how to change a Datafram

Answer 1

You can read the json data using

from pyspark import SQLContext

sqlContext = SQLContext(sc)
data_df = sqlContext.read.json("data.json", multiLine = True)

data_df.printSchema()

output

root
 |-- x: long (nullable = true)
 |-- y: struct (nullable = true)
 |    |-- p: struct (nullable = true)
 |    |    |-- name: string (nullable = true)
 |    |    |-- value: long (nullable = true)
 |    |-- q: struct (nullable = true)
 |    |    |-- name: string (nullable = true)
 |    |    |-- value: long (nullable = true)

Now you can access the data from y column as

data_df.select("y.p.name")
data_df.select("y.p.value")

output

abc, 10

Ok, the solution is to add a new nested column with correct schema and drop the column with wrong schema

from pyspark.sql.functions import *
from pyspark.sql import Row

df3 = spark.read.json("data.json", multiLine = True)

# create correct schema from old 
c = df3.schema['y'].jsonValue()
c['name'] = 'z'
c['type']['fields'][0]['type']['fields'][1]['type'] = 'long'
c['type']['fields'][1]['type']['fields'][1]['type'] = 'long'

y_schema = StructType.fromJson(c['type'])

# define a udf to populate the new column. Row are immuatable so you 
# have to build it from start.

def foo(row):
    d = Row.asDict(row)
    y = {}
    y["p"] = {}
    y["p"]["name"] = d["p"]["name"]
    y["p"]["value"] = int(d["p"]["value"])
    y["q"] = {}
    y["q"]["name"] = d["q"]["name"]
    y["q"]["value"] = int(d["p"]["value"])

    return(y)
map_foo = udf(foo, y_schema)

# add the column
df3_new  = df3.withColumn("z", map_foo("y"))

# delete the column
df4 = df3_new.drop("y")


df4.printSchema()

output

root
 |-- x: long (nullable = true)
 |-- z: struct (nullable = true)
 |    |-- p: struct (nullable = true)
 |    |    |-- name: string (nullable = true)
 |    |    |-- value: long (nullable = true)
 |    |-- q: struct (nullable = true)
 |    |    |-- name: string (nullable = true)
 |    |    |-- value: long (nullable = true)


df4.show()

output

+---+-------------------+
|  x|                  z|
+---+-------------------+
| 12|[[abc,10],[pqr,10]]|
+---+-------------------+