1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
|
// Copyright 2017 The TensorFlow Authors. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
// =============================================================================
#ifndef THIRD_PARTY_TENSORFLOW_CONTRIB_BOOSTED_TREES_LIB_UTILS_EXAMPLES_ITERABLE_H_
#define THIRD_PARTY_TENSORFLOW_CONTRIB_BOOSTED_TREES_LIB_UTILS_EXAMPLES_ITERABLE_H_
#include <vector>
#include "tensorflow/contrib/boosted_trees/lib/utils/example.h"
#include "tensorflow/contrib/boosted_trees/lib/utils/sparse_column_iterable.h"
#include "tensorflow/core/framework/tensor.h"
#include "tensorflow/core/util/sparse/sparse_tensor.h"
namespace tensorflow {
namespace boosted_trees {
namespace utils {
// Enables row-wise iteration through examples from feature columns.
class ExamplesIterable {
public:
// Constructs an iterable given the desired examples slice and corresponding
// feature columns.
ExamplesIterable(
const std::vector<Tensor>& dense_float_feature_columns,
const std::vector<sparse::SparseTensor>& sparse_float_feature_columns,
const std::vector<sparse::SparseTensor>& sparse_int_feature_columns,
int64 example_start, int64 example_end);
// Helper class to iterate through examples.
class Iterator {
public:
Iterator(ExamplesIterable* iter, int64 example_idx);
Iterator& operator++() {
// Advance to next example.
++example_idx_;
// Update sparse column iterables.
for (auto& it : sparse_float_column_iterators_) {
++it;
}
for (auto& it : sparse_int_column_iterators_) {
++it;
}
return (*this);
}
Iterator operator++(int) {
Iterator tmp(*this);
++(*this);
return tmp;
}
bool operator!=(const Iterator& other) const {
QCHECK_EQ(iter_, other.iter_);
return (example_idx_ != other.example_idx_);
}
bool operator==(const Iterator& other) const {
QCHECK_EQ(iter_, other.iter_);
return (example_idx_ == other.example_idx_);
}
const Example& operator*() {
// Set example index based on iterator.
example_.example_idx = example_idx_;
// Get dense float values per column.
auto& dense_float_features = example_.dense_float_features;
for (size_t dense_float_idx = 0;
dense_float_idx < dense_float_features.size(); ++dense_float_idx) {
dense_float_features[dense_float_idx] =
iter_->dense_float_column_values_[dense_float_idx](example_idx_, 0);
}
// Get sparse float values per column.
auto& sparse_float_features = example_.sparse_float_features;
// Iterate through each sparse float feature column.
for (size_t sparse_float_idx = 0;
sparse_float_idx < iter_->sparse_float_column_iterables_.size();
++sparse_float_idx) {
// Clear info from a previous instance.
sparse_float_features[sparse_float_idx].Clear();
// Get range for values tensor.
const auto& row_range =
(*sparse_float_column_iterators_[sparse_float_idx]);
DCHECK_EQ(example_idx_, row_range.example_idx);
// If the example has this feature column.
if (row_range.start < row_range.end) {
const int32 dimension =
iter_->sparse_float_dimensions_[sparse_float_idx];
sparse_float_features[sparse_float_idx].SetDimension(dimension);
if (dimension <= 1) {
// single dimensional sparse feature column.
DCHECK_EQ(1, row_range.end - row_range.start);
sparse_float_features[sparse_float_idx].Add(
0, iter_->sparse_float_column_values_[sparse_float_idx](
row_range.start));
} else {
// Retrieve original indices tensor.
const TTypes<int64>::ConstMatrix& indices =
iter_->sparse_float_column_iterables_[sparse_float_idx]
.sparse_indices();
sparse_float_features[sparse_float_idx].Reserve(row_range.end -
row_range.start);
// For each value.
for (int64 row_idx = row_range.start; row_idx < row_range.end;
++row_idx) {
// Get the feature id for the feature column and the value.
const int32 feature_id = indices(row_idx, 1);
DCHECK_EQ(example_idx_, indices(row_idx, 0));
// Save the value to our sparse matrix.
sparse_float_features[sparse_float_idx].Add(
feature_id,
iter_->sparse_float_column_values_[sparse_float_idx](
row_idx));
}
}
}
}
// Get sparse int values per column.
auto& sparse_int_features = example_.sparse_int_features;
for (size_t sparse_int_idx = 0;
sparse_int_idx < sparse_int_features.size(); ++sparse_int_idx) {
const auto& row_range = (*sparse_int_column_iterators_[sparse_int_idx]);
DCHECK_EQ(example_idx_, row_range.example_idx);
sparse_int_features[sparse_int_idx].clear();
if (row_range.start < row_range.end) {
sparse_int_features[sparse_int_idx].reserve(row_range.end -
row_range.start);
for (int64 row_idx = row_range.start; row_idx < row_range.end;
++row_idx) {
sparse_int_features[sparse_int_idx].insert(
iter_->sparse_int_column_values_[sparse_int_idx](row_idx));
}
}
}
return example_;
}
private:
// Examples iterable (not owned).
const ExamplesIterable* iter_;
// Example index.
int64 example_idx_;
// Sparse float column iterators.
std::vector<SparseColumnIterable::Iterator> sparse_float_column_iterators_;
// Sparse int column iterators.
std::vector<SparseColumnIterable::Iterator> sparse_int_column_iterators_;
// Example placeholder.
Example example_;
};
Iterator begin() { return Iterator(this, example_start_); }
Iterator end() { return Iterator(this, example_end_); }
private:
// Example slice spec.
const int64 example_start_;
const int64 example_end_;
// Dense float column values.
std::vector<TTypes<float>::ConstMatrix> dense_float_column_values_;
// Sparse float column iterables.
std::vector<SparseColumnIterable> sparse_float_column_iterables_;
// Sparse float column values.
std::vector<TTypes<float>::ConstVec> sparse_float_column_values_;
// Dimensions for sparse float feature columns.
std::vector<int32> sparse_float_dimensions_;
// Sparse int column iterables.
std::vector<SparseColumnIterable> sparse_int_column_iterables_;
// Sparse int column values.
std::vector<TTypes<int64>::ConstVec> sparse_int_column_values_;
};
} // namespace utils
} // namespace boosted_trees
} // namespace tensorflow
#endif // THIRD_PARTY_TENSORFLOW_CONTRIB_BOOSTED_TREES_LIB_UTILS_EXAMPLES_ITERABLE_H_
|
