Combining

Files
file	lists/combine.hpp

Enumerations
enum class	cudf::lists::concatenate_null_policy { IGNORE , NULLIFY_OUTPUT_ROW }
	Flag to specify whether a null list element will be ignored from concatenation, or the entire concatenation result involving null list elements will be a null element.

Functions
std::unique_ptr< column >	cudf::lists::concatenate_rows (table_view const &input, concatenate_null_policy null_policy=concatenate_null_policy::IGNORE, rmm::cuda_stream_view stream=cudf::get_default_stream(), rmm::mr::device_memory_resource *mr=rmm::mr::get_current_device_resource())
	Row-wise concatenating multiple lists columns into a single lists column. More...
std::unique_ptr< column >	cudf::lists::concatenate_list_elements (column_view const &input, concatenate_null_policy null_policy=concatenate_null_policy::IGNORE, rmm::cuda_stream_view stream=cudf::get_default_stream(), rmm::mr::device_memory_resource *mr=rmm::mr::get_current_device_resource())
	Concatenating multiple lists on the same row of a lists column into a single list. More...

Detailed Description

Function Documentation

concatenate_list_elements()

std::unique_ptr<column> cudf::lists::concatenate_list_elements	(	column_view const &	input,
		concatenate_null_policy	null_policy = concatenate_null_policy::IGNORE,
		rmm::cuda_stream_view	stream = cudf::get_default_stream(),
		rmm::mr::device_memory_resource *	mr = rmm::mr::get_current_device_resource()
	)

Concatenating multiple lists on the same row of a lists column into a single list.

Given a lists column where each row in the column is a list of lists of entries, an output lists column is generated by concatenating all the list elements at the same row together. If any row contains null list elements, the concatenation process will either ignore those null elements, or will simply set the entire resulting row to be a null element.

l = [ [{1, 2}, {3, 4}, {5}], [{6}, {}, {7, 8, 9}] ]
r = lists::concatenate_list_elements(l);
r is [ {1, 2, 3, 4, 5}, {6, 7, 8, 9} ]

Exceptions

std::invalid_argument

if the input column is not at least two-level depth lists column (i.e., each row must be a list of lists).

Parameters

input	The lists column containing lists of list elements to concatenate.
null_policy	The parameter to specify whether a null list element will be ignored from concatenation, or any concatenation involving a null element will result in a null list.
stream	CUDA stream used for device memory operations and kernel launches.
mr	Device memory resource used to allocate the returned column's device memory.

Returns

A new column in which each row is a list resulted from concatenating all list elements in the corresponding row of the input lists column.

concatenate_rows()

std::unique_ptr<column> cudf::lists::concatenate_rows	(	table_view const &	input,
		concatenate_null_policy	null_policy = concatenate_null_policy::IGNORE,
		rmm::cuda_stream_view	stream = cudf::get_default_stream(),
		rmm::mr::device_memory_resource *	mr = rmm::mr::get_current_device_resource()
	)

Row-wise concatenating multiple lists columns into a single lists column.

The output column is generated by concatenating the elements within each row of the input table. If any row of the input table contains null elements, the concatenation process will either ignore those null elements, or will simply set the entire resulting row to be a null element.

s1 = [{0, 1}, {2, 3, 4}, {5}, {}, {6, 7}]
s2 = [{8}, {9}, {}, {10, 11, 12}, {13, 14, 15, 16}]
r = lists::concatenate_rows(s1, s2)
r is now [{0, 1, 8}, {2, 3, 4, 9}, {5}, {10, 11, 12}, {6, 7, 13, 14, 15, 16}]

Exceptions

cudf::logic_error	if any column of the input table is not a lists column.
cudf::logic_error	if all lists columns do not have the same type.

Parameters

input	Table of lists to be concatenated.
null_policy	The parameter to specify whether a null list element will be ignored from concatenation, or any concatenation involving a null element will result in a null list.
stream	CUDA stream used for device memory operations and kernel launches.
mr	Device memory resource used to allocate the returned column's device memory.

Returns

A new column in which each row is a list resulted from concatenating all list elements in the corresponding row of the input table.