Lists Gather#
- group lists_gather
Functions
-
std::unique_ptr<column> segmented_gather(lists_column_view const &source_column, lists_column_view const &gather_map_list, out_of_bounds_policy bounds_policy = out_of_bounds_policy::DONT_CHECK, rmm::cuda_stream_view stream = cudf::get_default_stream(), rmm::mr::device_memory_resource *mr = rmm::mr::get_current_device_resource())#
Segmented gather of the elements within a list element in each row of a list column.
source_column
with any depth andgather_map_list
with depth 1 are only supported.source_column : [{"a", "b", "c", "d"}, {"1", "2", "3", "4"}, {"x", "y", "z"}] gather_map_list : [{0, 1, 3, 2}, {1, 3, 2}, {}] result : [{"a", "b", "d", "c"}, {"2", "4", "3"}, {}]
If indices in
gather_map_list
are outside the range[-n, n)
, wheren
is the number of elements in corresponding row of the source column, the behavior is as follows:If
bounds_policy
is set toDONT_CHECK
, the behavior is undefined.If
bounds_policy
is set toNULLIFY
, the corresponding element in the list row is set to null in the output column.
source_column : [{"a", "b", "c", "d"}, {"1", "2", "3", "4"}, {"x", "y", "z"}] gather_map_list : [{0, -1, 4, -5}, {1, 3, 5}, {}] result_with_nullify : [{"a", "d", null, null}, {"2", "4", null}, {}]
- Throws:
cudf::logic_error – if
gather_map_list
size is not same assource_column
size.std::invalid_argument – if gather_map contains null values.
cudf::logic_error – if gather_map is not list column of an index type.
- Parameters:
source_column – View into the list column to gather from
gather_map_list – View into a non-nullable list column of integral indices that maps the element in list of each row in the source columns to rows of lists in the destination columns.
bounds_policy – Can be
DONT_CHECK
orNULLIFY
. Selects whether or not to nullify the output list row’s element, when the gather index falls outside the range[-n, n)
, wheren
is the number of elements in list row corresponding to the gather-map row.stream – CUDA stream used for device memory operations and kernel launches.
mr – Device memory resource to allocate any returned objects
- Returns:
column with elements in list of rows gathered based on
gather_map_list
-
std::unique_ptr<column> segmented_gather(lists_column_view const &source_column, lists_column_view const &gather_map_list, out_of_bounds_policy bounds_policy = out_of_bounds_policy::DONT_CHECK, rmm::cuda_stream_view stream = cudf::get_default_stream(), rmm::mr::device_memory_resource *mr = rmm::mr::get_current_device_resource())#