view release on metacpan or  search on metacpan

xgboost/cub/cub/device/device_segmented_reduce.cuh  view on Meta::CPAN

/******************************************************************************
 * Copyright (c) 2011, Duane Merrill.  All rights reserved.
 * Copyright (c) 2011-2016, NVIDIA CORPORATION.  All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *     * Redistributions of source code must retain the above copyright
 *       notice, this list of conditions and the following disclaimer.
 *     * Redistributions in binary form must reproduce the above copyright
 *       notice, this list of conditions and the following disclaimer in the
 *       documentation and/or other materials provided with the distribution.
 *     * Neither the name of the NVIDIA CORPORATION nor the
 *       names of its contributors may be used to endorse or promote products
 *       derived from this software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
 * DISCLAIMED. IN NO EVENT SHALL NVIDIA CORPORATION BE LIABLE FOR ANY
 * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
 * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 *
 ******************************************************************************/

/**
 * \file
 * cub::DeviceSegmentedReduce provides device-wide, parallel operations for computing a batched reduction across multiple sequences of data items residing within device-accessible memory.
 */

#pragma once

#include <stdio.h>
#include <iterator>

#include "../iterator/arg_index_input_iterator.cuh"
#include "dispatch/dispatch_reduce.cuh"
#include "dispatch/dispatch_reduce_by_key.cuh"
#include "../util_type.cuh"
#include "../util_namespace.cuh"

/// Optional outer namespace(s)
CUB_NS_PREFIX

/// CUB namespace
namespace cub {


/**
 * \brief DeviceSegmentedReduce provides device-wide, parallel operations for computing a reduction across multiple sequences of data items residing within device-accessible memory. ![](reduce_logo.png)
 * \ingroup SegmentedModule
 *
 * \par Overview
 * A <a href="http://en.wikipedia.org/wiki/Reduce_(higher-order_function)"><em>reduction</em></a> (or <em>fold</em>)
 * uses a binary combining operator to compute a single aggregate from a sequence of input elements.
 *
 * \par Usage Considerations
 * \cdp_class{DeviceSegmentedReduce}
 *
 */
struct DeviceSegmentedReduce
{
    /**
     * \brief Computes a device-wide segmented reduction using the specified binary \p reduction_op functor.
     *
     * \par
     * - Does not support binary reduction operators that are non-commutative.
     * - When input a contiguous sequence of segments, a single sequence
     *   \p segment_offsets (of length <tt>num_segments+1</tt>) can be aliased
     *   for both the \p d_begin_offsets and \p d_end_offsets parameters (where
     *   the latter is specified as <tt>segment_offsets+1</tt>).
     * - \devicestorage
     *
     * \par Snippet
     * The code snippet below illustrates a custom min-reduction of a device vector of \p int data elements.
     * \par
     * \code
     * #include <cub/cub.cuh>   // or equivalently <cub/device/device_radix_sort.cuh>
     *
     * // CustomMin functor
     * struct CustomMin
     * {
     *     template <typename T>
     *     CUB_RUNTIME_FUNCTION __forceinline__
     *     T operator()(const T &a, const T &b) const {
     *         return (b < a) ? b : a;
     *     }
     * };
     *
     * // Declare, allocate, and initialize device-accessible pointers for input and output
     * int          num_segments;   // e.g., 3
     * int          *d_offsets;     // e.g., [0, 3, 3, 7]
     * int          *d_in;          // e.g., [8, 6, 7, 5, 3, 0, 9]
     * int          *d_out;         // e.g., [-, -, -]
     * CustomMin    min_op;
     * int          initial_value;           // e.g., INT_MAX
     * ...
     *
     * // Determine temporary device storage requirements
     * void     *d_temp_storage = NULL;
     * size_t   temp_storage_bytes = 0;
     * cub::DeviceSegmentedReduce::Reduce(d_temp_storage, temp_storage_bytes, d_in, d_out,
     *     num_segments, d_offsets, d_offsets + 1, min_op, initial_value);
     *
     * // Allocate temporary storage
     * cudaMalloc(&d_temp_storage, temp_storage_bytes);
     *
     * // Run reduction
     * cub::DeviceSegmentedReduce::Reduce(d_temp_storage, temp_storage_bytes, d_in, d_out,
     *     num_segments, d_offsets, d_offsets + 1, min_op, initial_value);