1// This file is part of Eigen, a lightweight C++ template library
2// for linear algebra.
3//
4// Copyright (C) 2016 Igor Babuschkin <igor@babuschk.in>
5//
6// This Source Code Form is subject to the terms of the Mozilla
7// Public License v. 2.0. If a copy of the MPL was not distributed
8// with this file, You can obtain one at http://mozilla.org/MPL/2.0/.
9
10#include "main.h"
11#include <limits>
12#include <numeric>
13#include <Eigen/CXX11/Tensor>
14
15using Eigen::Tensor;
16
17template <int DataLayout, typename Type=float, bool Exclusive = false>
18static void test_1d_scan()
19{
20  int size = 50;
21  Tensor<Type, 1, DataLayout> tensor(size);
22  tensor.setRandom();
23  Tensor<Type, 1, DataLayout> result = tensor.cumsum(0, Exclusive);
24
25  VERIFY_IS_EQUAL(tensor.dimension(0), result.dimension(0));
26
27  float accum = 0;
28  for (int i = 0; i < size; i++) {
29    if (Exclusive) {
30      VERIFY_IS_EQUAL(result(i), accum);
31      accum += tensor(i);
32    } else {
33      accum += tensor(i);
34      VERIFY_IS_EQUAL(result(i), accum);
35    }
36  }
37
38  accum = 1;
39  result = tensor.cumprod(0, Exclusive);
40  for (int i = 0; i < size; i++) {
41    if (Exclusive) {
42      VERIFY_IS_EQUAL(result(i), accum);
43      accum *= tensor(i);
44    } else {
45      accum *= tensor(i);
46      VERIFY_IS_EQUAL(result(i), accum);
47    }
48  }
49}
50
51template <int DataLayout, typename Type=float>
52static void test_4d_scan()
53{
54  int size = 5;
55  Tensor<Type, 4, DataLayout> tensor(size, size, size, size);
56  tensor.setRandom();
57
58  Tensor<Type, 4, DataLayout> result(size, size, size, size);
59
60  result = tensor.cumsum(0);
61  float accum = 0;
62  for (int i = 0; i < size; i++) {
63    accum += tensor(i, 1, 2, 3);
64    VERIFY_IS_EQUAL(result(i, 1, 2, 3), accum);
65  }
66  result = tensor.cumsum(1);
67  accum = 0;
68  for (int i = 0; i < size; i++) {
69    accum += tensor(1, i, 2, 3);
70    VERIFY_IS_EQUAL(result(1, i, 2, 3), accum);
71  }
72  result = tensor.cumsum(2);
73  accum = 0;
74  for (int i = 0; i < size; i++) {
75    accum += tensor(1, 2, i, 3);
76    VERIFY_IS_EQUAL(result(1, 2, i, 3), accum);
77  }
78  result = tensor.cumsum(3);
79  accum = 0;
80  for (int i = 0; i < size; i++) {
81    accum += tensor(1, 2, 3, i);
82    VERIFY_IS_EQUAL(result(1, 2, 3, i), accum);
83  }
84}
85
86template <int DataLayout>
87static void test_tensor_maps() {
88  int inputs[20];
89  TensorMap<Tensor<int, 1, DataLayout> > tensor_map(inputs, 20);
90  tensor_map.setRandom();
91
92  Tensor<int, 1, DataLayout> result = tensor_map.cumsum(0);
93
94  int accum = 0;
95  for (int i = 0; i < 20; ++i) {
96    accum += tensor_map(i);
97    VERIFY_IS_EQUAL(result(i), accum);
98  }
99}
100
101void test_cxx11_tensor_scan() {
102  CALL_SUBTEST((test_1d_scan<ColMajor, float, true>()));
103  CALL_SUBTEST((test_1d_scan<ColMajor, float, false>()));
104  CALL_SUBTEST((test_1d_scan<RowMajor, float, true>()));
105  CALL_SUBTEST((test_1d_scan<RowMajor, float, false>()));
106  CALL_SUBTEST(test_4d_scan<ColMajor>());
107  CALL_SUBTEST(test_4d_scan<RowMajor>());
108  CALL_SUBTEST(test_tensor_maps<ColMajor>());
109  CALL_SUBTEST(test_tensor_maps<RowMajor>());
110}
111