From d9978fb4e4d4de3a320b012373c18bd278462780 Mon Sep 17 00:00:00 2001 From: Jagadesh Kiran Date: Mon, 11 Mar 2019 08:53:09 -0500 Subject: [PATCH] [SPARK-26860][PYSPARK][SPARKR] Fix for RangeBetween and RowsBetween docs to be in sync with spark documentation The docs describing RangeBetween & RowsBetween for pySpark & SparkR are not in sync with Spark description. a. Edited PySpark and SparkR docs and made description same for both RangeBetween and RowsBetween b. created executable examples in both pySpark and SparkR documentation c. Locally tested the patch for scala Style checks and UT for checking no testcase failures Closes #23946 from jagadesh-kiran/master. Authored-by: Jagadesh Kiran Signed-off-by: Sean Owen --- R/pkg/R/WindowSpec.R | 41 ++++++++++++- python/pyspark/sql/window.py | 61 ++++++++++++++++++- .../apache/spark/sql/expressions/Window.scala | 4 +- 3 files changed, 101 insertions(+), 5 deletions(-) diff --git a/R/pkg/R/WindowSpec.R b/R/pkg/R/WindowSpec.R index ee7f4adf72..037809cd09 100644 --- a/R/pkg/R/WindowSpec.R +++ b/R/pkg/R/WindowSpec.R @@ -127,6 +127,16 @@ setMethod("orderBy", #' "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 \code{Window.unboundedPreceding}, \code{Window.unboundedFollowing}, +#' and \code{Window.currentRow} to specify special boundary values, rather than using long 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. +#' #' @param x a WindowSpec #' @param start boundary start, inclusive. #' The frame is unbounded if this is the minimum long value. @@ -139,7 +149,14 @@ setMethod("orderBy", #' @family windowspec_method #' @examples #' \dontrun{ -#' rowsBetween(ws, 0, 3) +#' id <- c(rep(1, 3), rep(2, 3), 3) +#' desc <- c('New', 'New', 'Good', 'New', 'Good', 'Good', 'New') +#' df <- data.frame(id, desc) +#' df <- createDataFrame(df) +#' w1 <- orderBy(windowPartitionBy('desc'), df$id) +#' w2 <- rowsBetween(w1, 0, 3) +#' df1 <- withColumn(df, "sum", over(sum(df$id), w2)) +#' head(df1) #' } #' @note rowsBetween since 2.0.0 setMethod("rowsBetween", @@ -158,6 +175,19 @@ setMethod("rowsBetween", #' "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 \code{Window.unboundedPreceding}, \code{Window.unboundedFollowing}, +#' and \code{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. +#' #' @param x a WindowSpec #' @param start boundary start, inclusive. #' The frame is unbounded if this is the minimum long value. @@ -170,7 +200,14 @@ setMethod("rowsBetween", #' @family windowspec_method #' @examples #' \dontrun{ -#' rangeBetween(ws, 0, 3) +#' id <- c(rep(1, 3), rep(2, 3), 3) +#' desc <- c('New', 'New', 'Good', 'New', 'Good', 'Good', 'New') +#' df <- data.frame(id, desc) +#' df <- createDataFrame(df) +#' w1 <- orderBy(windowPartitionBy('desc'), df$id) +#' w2 <- rangeBetween(w1, 0, 3) +#' df1 <- withColumn(df, "sum", over(sum(df$id), w2)) +#' head(df1) #' } #' @note rangeBetween since 2.0.0 setMethod("rangeBetween", diff --git a/python/pyspark/sql/window.py b/python/pyspark/sql/window.py index e76563dfaa..65c3ff5cec 100644 --- a/python/pyspark/sql/window.py +++ b/python/pyspark/sql/window.py @@ -97,6 +97,32 @@ class Window(object): 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. + + >>> from pyspark.sql import Window + >>> from pyspark.sql import functions as func + >>> from pyspark.sql import SQLContext + >>> sc = SparkContext.getOrCreate() + >>> sqlContext = SQLContext(sc) + >>> tup = [(1, "a"), (1, "a"), (2, "a"), (1, "b"), (2, "b"), (3, "b")] + >>> df = sqlContext.createDataFrame(tup, ["id", "category"]) + >>> window = Window.partitionBy("category").orderBy("id").rowsBetween(Window.currentRow, 1) + >>> df.withColumn("sum", func.sum("id").over(window)).show() + +---+--------+---+ + | id|category|sum| + +---+--------+---+ + | 1| b| 3| + | 2| b| 5| + | 3| b| 3| + | 1| a| 2| + | 1| a| 3| + | 2| a| 2| + +---+--------+---+ + :param start: boundary start, inclusive. The frame is unbounded if this is ``Window.unboundedPreceding``, or any value less than or equal to -9223372036854775808. @@ -127,6 +153,35 @@ class Window(object): and ``Window.currentRow`` to specify special boundary values, rather than using integral 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. + + >>> from pyspark.sql import Window + >>> from pyspark.sql import functions as func + >>> from pyspark.sql import SQLContext + >>> sc = SparkContext.getOrCreate() + >>> sqlContext = SQLContext(sc) + >>> tup = [(1, "a"), (1, "a"), (2, "a"), (1, "b"), (2, "b"), (3, "b")] + >>> df = sqlContext.createDataFrame(tup, ["id", "category"]) + >>> window = Window.partitionBy("category").orderBy("id").rangeBetween(Window.currentRow, 1) + >>> df.withColumn("sum", func.sum("id").over(window)).show() + +---+--------+---+ + | id|category|sum| + +---+--------+---+ + | 1| b| 3| + | 2| b| 5| + | 3| b| 3| + | 1| a| 4| + | 1| a| 4| + | 2| a| 2| + +---+--------+---+ + :param start: boundary start, inclusive. The frame is unbounded if this is ``Window.unboundedPreceding``, or any value less than or equal to max(-sys.maxsize, -9223372036854775808). @@ -231,8 +286,12 @@ class WindowSpec(object): def _test(): import doctest + import pyspark.sql.window SparkContext('local[4]', 'PythonTest') - (failure_count, test_count) = doctest.testmod() + globs = pyspark.sql.window.__dict__.copy() + (failure_count, test_count) = doctest.testmod( + pyspark.sql.window, globs=globs, + optionflags=doctest.NORMALIZE_WHITESPACE) if failure_count: sys.exit(-1) diff --git a/sql/core/src/main/scala/org/apache/spark/sql/expressions/Window.scala b/sql/core/src/main/scala/org/apache/spark/sql/expressions/Window.scala index 3d8d931af2..9a4ad44447 100644 --- a/sql/core/src/main/scala/org/apache/spark/sql/expressions/Window.scala +++ b/sql/core/src/main/scala/org/apache/spark/sql/expressions/Window.scala @@ -175,8 +175,8 @@ object Window { * 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 + * 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