org.apache.spark.sql.expressions

Class WindowSpec

Object

org.apache.spark.sql.expressions.WindowSpec

public class WindowSpec
extends Object

A window specification that defines the partitioning, ordering, and frame boundaries.

Use the static methods in Window to create a WindowSpec.

Since:

1.4.0

Method Summary

Modifier and Type	Method and Description
WindowSpec	orderBy(Column... cols) Defines the ordering columns in a WindowSpec.
WindowSpec	orderBy(scala.collection.Seq<Column> cols) Defines the ordering columns in a WindowSpec.
WindowSpec	orderBy(String colName, scala.collection.Seq<String> colNames) Defines the ordering columns in a WindowSpec.
WindowSpec	orderBy(String colName, String... colNames) Defines the ordering columns in a WindowSpec.
WindowSpec	partitionBy(Column... cols) Defines the partitioning columns in a WindowSpec.
WindowSpec	partitionBy(scala.collection.Seq<Column> cols) Defines the partitioning columns in a WindowSpec.
WindowSpec	partitionBy(String colName, scala.collection.Seq<String> colNames) Defines the partitioning columns in a WindowSpec.
WindowSpec	partitionBy(String colName, String... colNames) Defines the partitioning columns in a WindowSpec.
WindowSpec	rangeBetween(Column start, Column end) Defines the frame boundaries, from start (inclusive) to end (inclusive).
WindowSpec	rangeBetween(long start, long end) Defines the frame boundaries, from start (inclusive) to end (inclusive).
WindowSpec	rowsBetween(long start, long end) Defines the frame boundaries, from start (inclusive) to end (inclusive).

Methods inherited from class Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Detail

partitionBy

public WindowSpec partitionBy(String colName,
                              String... colNames)

Defines the partitioning columns in a WindowSpec.

Parameters:

colName - (undocumented)

colNames - (undocumented)

Returns:

(undocumented)

Since:

1.4.0

partitionBy

public WindowSpec partitionBy(Column... cols)

Defines the partitioning columns in a WindowSpec.

Parameters:

cols - (undocumented)

Returns:

(undocumented)

Since:

1.4.0

orderBy

public WindowSpec orderBy(String colName,
                          String... colNames)

Defines the ordering columns in a WindowSpec.

Parameters:

colName - (undocumented)

colNames - (undocumented)

Returns:

(undocumented)

Since:

1.4.0

orderBy

public WindowSpec orderBy(Column... cols)

Defines the ordering columns in a WindowSpec.

Parameters:

cols - (undocumented)

Returns:

(undocumented)

Since:

1.4.0

partitionBy

public WindowSpec partitionBy(String colName,
                              scala.collection.Seq<String> colNames)

Defines the partitioning columns in a WindowSpec.

Parameters:

colName - (undocumented)

colNames - (undocumented)

Returns:

(undocumented)

Since:

1.4.0

partitionBy

public WindowSpec partitionBy(scala.collection.Seq<Column> cols)

Defines the partitioning columns in a WindowSpec.

Parameters:

cols - (undocumented)

Returns:

(undocumented)

Since:

1.4.0

orderBy

public WindowSpec orderBy(String colName,
                          scala.collection.Seq<String> colNames)

Defines the ordering columns in a WindowSpec.

Parameters:

colName - (undocumented)

colNames - (undocumented)

Returns:

(undocumented)

Since:

1.4.0

orderBy

public WindowSpec orderBy(scala.collection.Seq<Column> cols)

Defines the ordering columns in a WindowSpec.

Parameters:

cols - (undocumented)

Returns:

(undocumented)

Since:

1.4.0

rowsBetween

public WindowSpec rowsBetween(long start,
                              long end)

Defines the frame boundaries, from start (inclusive) to end (inclusive).

Both start and end are relative positions from the current row. For example, "0" means "current row", while "-1" means the row before the current row, and "5" means the fifth row after the current row.

We recommend users use Window.unboundedPreceding, Window.unboundedFollowing, and Window.currentRow to specify special boundary values, rather than using integral values directly.

A row based boundary is based on the position of the row within the partition. An offset indicates the number of rows above or below the current row, the frame for the current row starts or ends. For instance, given a row based sliding frame with a lower bound offset of -1 and a upper bound offset of +2. The frame for row with index 5 would range from index 4 to index 6.

   import org.apache.spark.sql.expressions.Window
   val df = Seq((1, "a"), (1, "a"), (2, "a"), (1, "b"), (2, "b"), (3, "b"))
     .toDF("id", "category")
   val byCategoryOrderedById =
     Window.partitionBy('category).orderBy('id).rowsBetween(Window.currentRow, 1)
   df.withColumn("sum", sum('id) over byCategoryOrderedById).show()

   +---+--------+---+
   | id|category|sum|
   +---+--------+---+
   |  1|       b|  3|
   |  2|       b|  5|
   |  3|       b|  3|
   |  1|       a|  2|
   |  1|       a|  3|
   |  2|       a|  2|
   +---+--------+---+
 

Parameters:

start - boundary start, inclusive. The frame is unbounded if this is the minimum long value (Window.unboundedPreceding).

end - boundary end, inclusive. The frame is unbounded if this is the maximum long value (Window.unboundedFollowing).

Returns:

(undocumented)

Since:

1.4.0

rangeBetween

public WindowSpec rangeBetween(long start,
                               long end)

Defines the frame boundaries, from start (inclusive) to end (inclusive).

Both start and end are relative from the current row. For example, "0" means "current row", while "-1" means one off before the current row, and "5" means the five off after the current row.

We recommend users use Window.unboundedPreceding, Window.unboundedFollowing, and Window.currentRow to specify special boundary values, rather than using long values directly.

A range-based boundary is based on the actual value of the ORDER BY expression(s). An offset is used to alter the value of the ORDER BY expression, for instance if the current order by expression has a value of 10 and the lower bound offset is -3, the resulting lower bound for the current row will be 10 - 3 = 7. This however puts a number of constraints on the ORDER BY expressions: there can be only one expression and this expression must have a numerical data type. An exception can be made when the offset is unbounded, because no value modification is needed, in this case multiple and non-numeric ORDER BY expression are allowed.

   import org.apache.spark.sql.expressions.Window
   val df = Seq((1, "a"), (1, "a"), (2, "a"), (1, "b"), (2, "b"), (3, "b"))
     .toDF("id", "category")
   val byCategoryOrderedById =
     Window.partitionBy('category).orderBy('id).rangeBetween(Window.currentRow, 1)
   df.withColumn("sum", sum('id) over byCategoryOrderedById).show()

   +---+--------+---+
   | id|category|sum|
   +---+--------+---+
   |  1|       b|  3|
   |  2|       b|  5|
   |  3|       b|  3|
   |  1|       a|  4|
   |  1|       a|  4|
   |  2|       a|  2|
   +---+--------+---+
 

Parameters:

start - boundary start, inclusive. The frame is unbounded if this is the minimum long value (Window.unboundedPreceding).

end - boundary end, inclusive. The frame is unbounded if this is the maximum long value (Window.unboundedFollowing).

Returns:

(undocumented)

Since:

1.4.0

rangeBetween

public WindowSpec rangeBetween(Column start,
                               Column end)

Defines the frame boundaries, from start (inclusive) to end (inclusive).

Both start and end are relative to the current row. For example, "lit(0)" means "current row", while "lit(-1)" means one off before the current row, and "lit(5)" means the five off after the current row.

Users should use unboundedPreceding(), unboundedFollowing(), and currentRow() from functions to specify special boundary values, literals are not transformed to SpecialFrameBoundarys.

A range-based boundary is based on the actual value of the ORDER BY expression(s). An offset is used to alter the value of the ORDER BY expression, for instance if the current order by expression has a value of 10 and the lower bound offset is -3, the resulting lower bound for the current row will be 10 - 3 = 7. This however puts a number of constraints on the ORDER BY expressions: there can be only one expression and this expression must have a numerical/date/timestamp data type. An exception can be made when the offset is unbounded, because no value modification is needed, in this case multiple and non-numerical/date/timestamp data type ORDER BY expression are allowed.

   import org.apache.spark.sql.expressions.Window
   val df = Seq((1, "a"), (1, "a"), (2, "a"), (1, "b"), (2, "b"), (3, "b"))
     .toDF("id", "category")
   val byCategoryOrderedById =
     Window.partitionBy('category).orderBy('id).rangeBetween(currentRow(), lit(1))
   df.withColumn("sum", sum('id) over byCategoryOrderedById).show()

   +---+--------+---+
   | id|category|sum|
   +---+--------+---+
   |  1|       b|  3|
   |  2|       b|  5|
   |  3|       b|  3|
   |  1|       a|  4|
   |  1|       a|  4|
   |  2|       a|  2|
   +---+--------+---+
 

Parameters:

start - boundary start, inclusive. The frame is unbounded if the expression is UnboundedPreceding.

end - boundary end, inclusive. The frame is unbounded if the expression is UnboundedFollowing.

Returns:

(undocumented)

Since:

2.3.0