Defining DataFrame Schemas with StructField and StructType

Spark DataFrames schemas are defined as a collection of typed columns. The entire schema is stored as a StructType and individual columns are stored as StructFields.

This blog post explains how to create and modify Spark schemas via the StructType and StructField classes.

We’ll show how to work with IntegerType, StringType, LongType, ArrayType, MapType and StructType columns.

Mastering Spark schemas is necessary for debugging code and writing tests.

This blog post provides a great introduction to these topics, but Writing Beautiful Spark Code provides a much more comprehensive review of the topics covered in this post. The book is the fastest way for you to become a strong Spark programmer.

Defining a schema to create a DataFrame

Let’s invent some sample data, define a schema, and create a DataFrame.

import org.apache.spark.sql.types._

val data = Seq(
  Row(8, "bat"),
  Row(64, "mouse"),
  Row(-27, "horse")
)

val schema = StructType(
  List(
    StructField("number", IntegerType, true),
    StructField("word", StringType, true)
  )
)

val df = spark.createDataFrame(
  spark.sparkContext.parallelize(data),
  schema
)
df.show()

+------+-----+
|number| word|
+------+-----+
|     8|  bat|
|    64|mouse|
|   -27|horse|
+------+-----+

StructType objects are instantiated with a List of StructField objects.

The org.apache.spark.sql.types package must be imported to access StructType, StructField, IntegerType, and StringType.

The createDataFrame() method takes two arguments:

  1. RDD of the data
  2. The DataFrame schema (a StructType object)

The schema() method returns a StructType object:

df.schema

StructType(
  StructField(number,IntegerType,true),
  StructField(word,StringType,true)
)

StructField

StructFields model each column in a DataFrame.

StructField objects are created with the name, dataType, and nullable properties. Here’s an example:

StructField("word", StringType, true)

The StructField above sets the name field to "word", the dataType field to StringType, and the nullable field to true.

"word" is the name of the column in the DataFrame.

StringType means that the column can only take string values like "hello" – it cannot take other values like 34 or false.

When the nullable field is set to true, the column can accept null values.

Defining schemas with the :: operator

We can also define a schema with the :: operator, like the examples in the StructType documentation.

val schema = StructType(
  StructField("number", IntegerType, true) ::
  StructField("word", StringType, true) :: Nil
)

The :: operator makes it easy to construct lists in Scala. We can also use :: to make a list of numbers.

5 :: 4 :: Nil

Notice that the last element always has to be Nil or the code will error out.

Defining schemas with the add() method

We can use the StructType#add() method to define schemas.

val schema = StructType(Seq(StructField("number", IntegerType, true)))
  .add(StructField("word", StringType, true))

add() is an overloaded method and there are several different ways to invoke it – this will work too:

val schema = StructType(Seq(StructField("number", IntegerType, true)))
  .add("word", StringType, true)

Check the StructType documentation for all the different ways add() can be used.

Common errors

Extra column defined in Schema

The following code has an extra column defined in the schema and will error out with this message: java.lang.RuntimeException: Error while encoding: java.lang.ArrayIndexOutOfBoundsException: 2.

val data = Seq(
  Row(8, "bat"),
  Row(64, "mouse"),
  Row(-27, "horse")
)

val schema = StructType(
  List(
    StructField("number", IntegerType, true),
    StructField("word", StringType, true),
    StructField("num2", IntegerType, true)
  )
)

val df = spark.createDataFrame(
  spark.sparkContext.parallelize(data),
  schema
)

The data only contains two columns, but the schema contains three StructField columns.

Type mismatch

The following code incorrectly characterizes a string column as an integer column and will error out with this message: java.lang.RuntimeException: Error while encoding: java.lang.RuntimeException: java.lang.String is not a valid external type for schema of int.

val data = Seq(
  Row(8, "bat"),
  Row(64, "mouse"),
  Row(-27, "horse")
)

val schema = StructType(
  List(
    StructField("num1", IntegerType, true),
    StructField("num2", IntegerType, true)
  )
)

val df = spark.createDataFrame(
  spark.sparkContext.parallelize(data),
  schema
)

df.show()

The first column of data (8, 64, and -27) can be characterized as IntegerType data.

The second column of data ("bat", "mouse", and "horse") cannot be characterized as an IntegerType column – this could would work if this column was recharacterized as StringType.

Nullable property exception

The following code incorrectly tries to add null to a column with a nullable property set to false and will error out with this message: java.lang.RuntimeException: Error while encoding: java.lang.RuntimeException: The 0th field 'word1' of input row cannot be null.

val data = Seq(
  Row("hi", "bat"),
  Row("bye", "mouse"),
  Row(null, "horse")
)

val schema = StructType(
  List(
    StructField("word1", StringType, false),
    StructField("word2", StringType, true)
  )
)

val df = spark.createDataFrame(
  spark.sparkContext.parallelize(data),
  schema
)

df.show()

LongType

Integers use 32 bits whereas long values use 64 bits.

Integers can hold values between -2 billion to 2 billion (-scala.math.pow(2, 31) to scala.math.pow(2, 31) - 1 to be exact).

Long values are suitable for bigger integers. You can create a long value in Scala by appending L to an integer – e.g. 4L or -60L.

Let’s create a DataFrame with a LongType column.

val data = Seq(
  Row(5L, "bat"),
  Row(-10L, "mouse"),
  Row(4L, "horse")
)

val schema = StructType(
  List(
    StructField("long_num", LongType, true),
    StructField("word", StringType, true)
  )
)

val df = spark.createDataFrame(
  spark.sparkContext.parallelize(data),
  schema
)
df.show()

+--------+-----+
|long_num| word|
+--------+-----+
|       5|  bat|
|     -10|mouse|
|       4|horse|
+--------+-----+

You’ll get the following error message if you try to add integers to a LongType column: java.lang.RuntimeException: Error while encoding: java.lang.RuntimeException: java.lang.Integer is not a valid external type for schema of bigint

Here’s an example of the erroneous code:

val data = Seq(
  Row(45, "bat"),
  Row(2, "mouse"),
  Row(3, "horse")
)

val schema = StructType(
  List(
    StructField("long_num", LongType, true),
    StructField("word", StringType, true)
  )
)

val df = spark.createDataFrame(
  spark.sparkContext.parallelize(data),
  schema
)

df.show()

ArrayType

Spark supports columns that contain arrays of values. Scala offers lists, sequences, and arrays. In regular Scala code, it’s best to use List or Seq, but Arrays are frequently used with Spark.

Here’s how to create an array of numbers with Scala:

val numbers = Array(1, 2, 3)

Let’s create a DataFrame with an ArrayType column.

val data = Seq(
  Row("bieber", Array("baby", "sorry")),
  Row("ozuna", Array("criminal"))
)

val schema = StructType(
  List(
    StructField("name", StringType, true),
    StructField("hit_songs", ArrayType(StringType, true), true)
  )
)

val df = spark.createDataFrame(
  spark.sparkContext.parallelize(data),
  schema
)
df.show()

+------+-------------+
|  name|    hit_songs|
+------+-------------+
|bieber|[baby, sorry]|
| ozuna|   [criminal]|
+------+-------------+

MapType

Scala maps store key / value pairs (maps are called “hashes” in other programming languages). Let’s create a Scala map with beers and their country of origin.

val beers = Map("aguila" -> "Colombia", "modelo" -> "Mexico")

Let’s grab the value that’s associated with the key "modelo":

beers("modelo") // Mexico

Let’s create a DataFrame with a MapType column.

val data = Seq(
  Row("sublime", Map(
    "good_song" -> "santeria",
    "bad_song" -> "doesn't exist")
  ),
  Row("prince_royce", Map(
    "good_song" -> "darte un beso",
    "bad_song" -> "back it up")
  )
)

val schema = StructType(
  List(
    StructField("name", StringType, true),
    StructField("songs", MapType(StringType, StringType, true), true)
  )
)

val df = spark.createDataFrame(
  spark.sparkContext.parallelize(data),
  schema
)
df.show(false)

+------------+----------------------------------------------------+
|name        |songs                                               |
+------------+----------------------------------------------------+
|sublime     |[good_song -> santeria, bad_song -> doesn't exist]  |
|prince_royce|[good_song -> darte un beso, bad_song -> back it up]|
+------------+----------------------------------------------------+

Notice that MapType is instantiated with three arguments (e.g. MapType(StringType, StringType, true)). The first argument is the keyType, the second argument is the valueType, and the third argument is a boolean flag for valueContainsNull. Map values can contain null if valueContainsNull is set to true, but the key can never be null.

StructType nested schemas

DataFrame schemas can be nested. A DataFrame column can be a struct – it’s essentially a schema within a schema.

Let’s create a DataFrame with a StructType column.

val data = Seq(
  Row("bob", Row("blue", 45)),
  Row("mary", Row("red", 64))
)

val schema = StructType(
  List(
    StructField("name", StringType, true),
    StructField(
      "person_details",
      StructType(
        List(
          StructField("favorite_color", StringType, true),
          StructField("age", IntegerType, true)
        )
      ),
      true
    )
  )
)

val df = spark.createDataFrame(
  spark.sparkContext.parallelize(data),
  schema
)
df.show()

+----+--------------+
|name|person_details|
+----+--------------+
| bob|    [blue, 45]|
|mary|     [red, 64]|
+----+--------------+

We can use the printSchema() method to illustrate that person_details is a struct column:

df.printSchema()

root
 |-- name: string (nullable = true)
 |-- person_details: struct (nullable = true)
 |    |-- favorite_color: string (nullable = true)
 |    |-- age: integer (nullable = true)

StructType object oriented programming

The StructType object mixes in the Seq trait to access a bunch of collection methods.

Here’s how StructType is defined:

case class StructType(fields: Array[StructField])
  extends DataType
  with Seq[StructField]

Here’s the StructType source code.

The Scala Seq trait is defined as follows:

trait Seq[+A]
  extends PartialFunction[Int, A]
  with Iterable[A]
  with GenSeq[A]
  with GenericTraversableTemplate[A, Seq]
  with SeqLike[A, Seq[A]]

By inheriting from the Seq trait, the StructType class gets access to collection methods like collect() and foldLeft().

Let’s create a DataFrame schema and use the foldLeft() method to create a sequence of all the column names.

val data = Seq(
  Row(8, "bat"),
  Row(64, "mouse"),
  Row(-27, "horse")
)

val schema = StructType(
  List(
    StructField("number", IntegerType, true),
    StructField("word", StringType, true)
  )
)

val df = spark.createDataFrame(
  spark.sparkContext.parallelize(data),
  schema
)

val columns = schema.foldLeft(Seq.empty[String]) {(memo: Seq[String], s: StructField) =>
  memo ++ Seq(s.name)
}

If we really wanted to get a list of all the column names, we could just run df.columns, but the foldLeft() method is clearly more powerful – it let’s us perform arbitrary collection operations on our DataFrame schemas.

Flattening DataFrames with StructType columns

In the previous section, we created a DataFrame with a StructType column. Let’s expand the two columns in the nested StructType column to be two separate fields.

We will leverage a flattenSchema method from spark-daria to make this easy.

import com.github.mrpowers.spark.daria.sql.DataFrameExt._

val flattenedDF = df.flattenSchema(delimiter = "_")
flattenedDF.show()

+----+-----------------------------+------------------+
|name|person_details_favorite_color|person_details_age|
+----+-----------------------------+------------------+
| bob|                         blue|                45|
|mary|                          red|                64|
+----+-----------------------------+------------------+

Take a look at the spark-daria source code to see how this code works.

Next steps

You’ll be defining a lot of schemas in your test suites so make sure to master all the concepts covered in this blog post.

Registration

1 Comment


  1. Hi, how can I create a data frame with columns of DateTime and Double types? Couldn’t find the answer anywhere.

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *