-
Notifications
You must be signed in to change notification settings - Fork 3.8k
/
dense_nbits_bin.hpp
397 lines (348 loc) · 13.8 KB
/
dense_nbits_bin.hpp
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
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
#ifndef LIGHTGBM_IO_DENSE_NBITS_BIN_HPP_
#define LIGHTGBM_IO_DENSE_NBITS_BIN_HPP_
#include <LightGBM/bin.h>
#include <vector>
#include <cstring>
#include <cstdint>
namespace LightGBM {
class Dense4bitsBin;
class Dense4bitsBinIterator : public BinIterator {
public:
explicit Dense4bitsBinIterator(const Dense4bitsBin* bin_data, uint32_t min_bin, uint32_t max_bin, uint32_t default_bin)
: bin_data_(bin_data), min_bin_(static_cast<uint8_t>(min_bin)),
max_bin_(static_cast<uint8_t>(max_bin)),
default_bin_(static_cast<uint8_t>(default_bin)) {
if (default_bin_ == 0) {
bias_ = 1;
} else {
bias_ = 0;
}
}
inline uint32_t RawGet(data_size_t idx) override;
inline uint32_t Get(data_size_t idx) override;
inline void Reset(data_size_t) override {}
private:
const Dense4bitsBin* bin_data_;
uint8_t min_bin_;
uint8_t max_bin_;
uint8_t default_bin_;
uint8_t bias_;
};
class Dense4bitsBin : public Bin {
public:
friend Dense4bitsBinIterator;
Dense4bitsBin(data_size_t num_data)
: num_data_(num_data) {
int len = (num_data_ + 1) / 2;
data_ = std::vector<uint8_t>(len, static_cast<uint8_t>(0));
buf_ = std::vector<uint8_t>(len, static_cast<uint8_t>(0));
}
~Dense4bitsBin() {
}
void Push(int, data_size_t idx, uint32_t value) override {
const int i1 = idx >> 1;
const int i2 = (idx & 1) << 2;
const uint8_t val = static_cast<uint8_t>(value) << i2;
if (i2 == 0) {
data_[i1] = val;
} else {
buf_[i1] = val;
}
}
void ReSize(data_size_t num_data) override {
if (num_data_ != num_data) {
num_data_ = num_data;
const int len = (num_data_ + 1) / 2;
data_.resize(len);
}
}
inline BinIterator* GetIterator(uint32_t min_bin, uint32_t max_bin, uint32_t default_bin) const override;
void ConstructHistogram(const data_size_t* data_indices, data_size_t num_data,
const score_t* ordered_gradients, const score_t* ordered_hessians,
HistogramBinEntry* out) const override {
const data_size_t rest = num_data & 0x3;
data_size_t i = 0;
for (; i < num_data - rest; i += 4) {
const data_size_t idx0 = data_indices[i];
const auto bin0 = (data_[idx0 >> 1] >> ((idx0 & 1) << 2)) & 0xf;
const data_size_t idx1 = data_indices[i + 1];
const auto bin1 = (data_[idx1 >> 1] >> ((idx1 & 1) << 2)) & 0xf;
const data_size_t idx2 = data_indices[i + 2];
const auto bin2 = (data_[idx2 >> 1] >> ((idx2 & 1) << 2)) & 0xf;
const data_size_t idx3 = data_indices[i + 3];
const auto bin3 = (data_[idx3 >> 1] >> ((idx3 & 1) << 2)) & 0xf;
out[bin0].sum_gradients += ordered_gradients[i];
out[bin1].sum_gradients += ordered_gradients[i + 1];
out[bin2].sum_gradients += ordered_gradients[i + 2];
out[bin3].sum_gradients += ordered_gradients[i + 3];
out[bin0].sum_hessians += ordered_hessians[i];
out[bin1].sum_hessians += ordered_hessians[i + 1];
out[bin2].sum_hessians += ordered_hessians[i + 2];
out[bin3].sum_hessians += ordered_hessians[i + 3];
++out[bin0].cnt;
++out[bin1].cnt;
++out[bin2].cnt;
++out[bin3].cnt;
}
for (; i < num_data; ++i) {
const data_size_t idx = data_indices[i];
const auto bin = (data_[idx >> 1] >> ((idx & 1) << 2)) & 0xf;
out[bin].sum_gradients += ordered_gradients[i];
out[bin].sum_hessians += ordered_hessians[i];
++out[bin].cnt;
}
}
void ConstructHistogram(data_size_t num_data,
const score_t* ordered_gradients, const score_t* ordered_hessians,
HistogramBinEntry* out) const override {
const data_size_t rest = num_data & 0x3;
data_size_t i = 0;
for (; i < num_data - rest; i += 4) {
const auto bin0 = (data_[i >> 1]) & 0xf;
const auto bin1 = (data_[i >> 1] >> 4) & 0xf;
const auto bin2 = (data_[(i >> 1) + 1]) & 0xf;
const auto bin3 = (data_[(i >> 1) + 1] >> 4) & 0xf;
out[bin0].sum_gradients += ordered_gradients[i];
out[bin1].sum_gradients += ordered_gradients[i + 1];
out[bin2].sum_gradients += ordered_gradients[i + 2];
out[bin3].sum_gradients += ordered_gradients[i + 3];
out[bin0].sum_hessians += ordered_hessians[i];
out[bin1].sum_hessians += ordered_hessians[i + 1];
out[bin2].sum_hessians += ordered_hessians[i + 2];
out[bin3].sum_hessians += ordered_hessians[i + 3];
++out[bin0].cnt;
++out[bin1].cnt;
++out[bin2].cnt;
++out[bin3].cnt;
}
for (; i < num_data; ++i) {
const auto bin = (data_[i >> 1] >> ((i & 1) << 2)) & 0xf;
out[bin].sum_gradients += ordered_gradients[i];
out[bin].sum_hessians += ordered_hessians[i];
++out[bin].cnt;
}
}
void ConstructHistogram(const data_size_t* data_indices, data_size_t num_data,
const score_t* ordered_gradients,
HistogramBinEntry* out) const override {
const data_size_t rest = num_data & 0x3;
data_size_t i = 0;
for (; i < num_data - rest; i += 4) {
const data_size_t idx0 = data_indices[i];
const auto bin0 = (data_[idx0 >> 1] >> ((idx0 & 1) << 2)) & 0xf;
const data_size_t idx1 = data_indices[i + 1];
const auto bin1 = (data_[idx1 >> 1] >> ((idx1 & 1) << 2)) & 0xf;
const data_size_t idx2 = data_indices[i + 2];
const auto bin2 = (data_[idx2 >> 1] >> ((idx2 & 1) << 2)) & 0xf;
const data_size_t idx3 = data_indices[i + 3];
const auto bin3 = (data_[idx3 >> 1] >> ((idx3 & 1) << 2)) & 0xf;
out[bin0].sum_gradients += ordered_gradients[i];
out[bin1].sum_gradients += ordered_gradients[i + 1];
out[bin2].sum_gradients += ordered_gradients[i + 2];
out[bin3].sum_gradients += ordered_gradients[i + 3];
++out[bin0].cnt;
++out[bin1].cnt;
++out[bin2].cnt;
++out[bin3].cnt;
}
for (; i < num_data; ++i) {
const data_size_t idx = data_indices[i];
const auto bin = (data_[idx >> 1] >> ((idx & 1) << 2)) & 0xf;
out[bin].sum_gradients += ordered_gradients[i];
++out[bin].cnt;
}
}
void ConstructHistogram(data_size_t num_data,
const score_t* ordered_gradients,
HistogramBinEntry* out) const override {
const data_size_t rest = num_data & 0x3;
data_size_t i = 0;
for (; i < num_data - rest; i += 4) {
const auto bin0 = (data_[i >> 1]) & 0xf;
const auto bin1 = (data_[i >> 1] >> 4) & 0xf;
const auto bin2 = (data_[(i >> 1) + 1]) & 0xf;
const auto bin3 = (data_[(i >> 1) + 1] >> 4) & 0xf;
out[bin0].sum_gradients += ordered_gradients[i];
out[bin1].sum_gradients += ordered_gradients[i + 1];
out[bin2].sum_gradients += ordered_gradients[i + 2];
out[bin3].sum_gradients += ordered_gradients[i + 3];
++out[bin0].cnt;
++out[bin1].cnt;
++out[bin2].cnt;
++out[bin3].cnt;
}
for (; i < num_data; ++i) {
const auto bin = (data_[i >> 1] >> ((i & 1) << 2)) & 0xf;
out[bin].sum_gradients += ordered_gradients[i];
++out[bin].cnt;
}
}
virtual data_size_t Split(
uint32_t min_bin, uint32_t max_bin, uint32_t default_bin, MissingType missing_type, bool default_left,
uint32_t threshold, data_size_t* data_indices, data_size_t num_data,
data_size_t* lte_indices, data_size_t* gt_indices) const override {
if (num_data <= 0) { return 0; }
uint8_t th = static_cast<uint8_t>(threshold + min_bin);
const uint8_t minb = static_cast<uint8_t>(min_bin);
const uint8_t maxb = static_cast<uint8_t>(max_bin);
uint8_t t_default_bin = static_cast<uint8_t>(min_bin + default_bin);
if (default_bin == 0) {
th -= 1;
t_default_bin -= 1;
}
data_size_t lte_count = 0;
data_size_t gt_count = 0;
data_size_t* default_indices = gt_indices;
data_size_t* default_count = >_count;
if (missing_type == MissingType::NaN) {
if (default_bin <= threshold) {
default_indices = lte_indices;
default_count = <e_count;
}
data_size_t* missing_default_indices = gt_indices;
data_size_t* missing_default_count = >_count;
if (default_left) {
missing_default_indices = lte_indices;
missing_default_count = <e_count;
}
for (data_size_t i = 0; i < num_data; ++i) {
const data_size_t idx = data_indices[i];
const uint8_t bin = (data_[idx >> 1] >> ((idx & 1) << 2)) & 0xf;
if (bin < minb || bin > maxb || t_default_bin == bin) {
default_indices[(*default_count)++] = idx;
} else if (bin == maxb) {
missing_default_indices[(*missing_default_count)++] = idx;
} else if (bin > th) {
gt_indices[gt_count++] = idx;
} else {
lte_indices[lte_count++] = idx;
}
}
} else {
if ((default_left && missing_type == MissingType::Zero) || (default_bin <= threshold && missing_type != MissingType::Zero)) {
default_indices = lte_indices;
default_count = <e_count;
}
for (data_size_t i = 0; i < num_data; ++i) {
const data_size_t idx = data_indices[i];
const uint8_t bin = (data_[idx >> 1] >> ((idx & 1) << 2)) & 0xf;
if (bin < minb || bin > maxb || t_default_bin == bin) {
default_indices[(*default_count)++] = idx;
} else if (bin > th) {
gt_indices[gt_count++] = idx;
} else {
lte_indices[lte_count++] = idx;
}
}
}
return lte_count;
}
virtual data_size_t SplitCategorical(
uint32_t min_bin, uint32_t max_bin, uint32_t default_bin,
const uint32_t* threshold, int num_threahold, data_size_t* data_indices, data_size_t num_data,
data_size_t* lte_indices, data_size_t* gt_indices) const override {
if (num_data <= 0) { return 0; }
data_size_t lte_count = 0;
data_size_t gt_count = 0;
data_size_t* default_indices = gt_indices;
data_size_t* default_count = >_count;
if (Common::FindInBitset(threshold, num_threahold, default_bin)) {
default_indices = lte_indices;
default_count = <e_count;
}
for (data_size_t i = 0; i < num_data; ++i) {
const data_size_t idx = data_indices[i];
const uint32_t bin = (data_[idx >> 1] >> ((idx & 1) << 2)) & 0xf;
if (bin < min_bin || bin > max_bin) {
default_indices[(*default_count)++] = idx;
} else if (Common::FindInBitset(threshold, num_threahold, bin - min_bin)) {
lte_indices[lte_count++] = idx;
} else {
gt_indices[gt_count++] = idx;
}
}
return lte_count;
}
data_size_t num_data() const override { return num_data_; }
/*! \brief not ordered bin for dense feature */
OrderedBin* CreateOrderedBin() const override { return nullptr; }
void FinishLoad() override {
if (buf_.empty()) { return; }
int len = (num_data_ + 1) / 2;
for (int i = 0; i < len; ++i) {
data_[i] |= buf_[i];
}
buf_.clear();
}
void LoadFromMemory(const void* memory, const std::vector<data_size_t>& local_used_indices) override {
const uint8_t* mem_data = reinterpret_cast<const uint8_t*>(memory);
if (!local_used_indices.empty()) {
const data_size_t rest = num_data_ & 1;
for (int i = 0; i < num_data_ - rest; i += 2) {
// get old bins
data_size_t idx = local_used_indices[i];
const auto bin1 = static_cast<uint8_t>((mem_data[idx >> 1] >> ((idx & 1) << 2)) & 0xf);
idx = local_used_indices[i + 1];
const auto bin2 = static_cast<uint8_t>((mem_data[idx >> 1] >> ((idx & 1) << 2)) & 0xf);
// add
const int i1 = i >> 1;
data_[i1] = (bin1 | (bin2 << 4));
}
if (rest) {
data_size_t idx = local_used_indices[num_data_ - 1];
data_[num_data_ >> 1] = (mem_data[idx >> 1] >> ((idx & 1) << 2)) & 0xf;
}
} else {
for (size_t i = 0; i < data_.size(); ++i) {
data_[i] = mem_data[i];
}
}
}
void CopySubset(const Bin* full_bin, const data_size_t* used_indices, data_size_t num_used_indices) override {
auto other_bin = dynamic_cast<const Dense4bitsBin*>(full_bin);
const data_size_t rest = num_used_indices & 1;
for (int i = 0; i < num_used_indices - rest; i += 2) {
data_size_t idx = used_indices[i];
const auto bin1 = static_cast<uint8_t>((other_bin->data_[idx >> 1] >> ((idx & 1) << 2)) & 0xf);
idx = used_indices[i + 1];
const auto bin2 = static_cast<uint8_t>((other_bin->data_[idx >> 1] >> ((idx & 1) << 2)) & 0xf);
const int i1 = i >> 1;
data_[i1] = (bin1 | (bin2 << 4));
}
if (rest) {
data_size_t idx = used_indices[num_used_indices - 1];
data_[num_used_indices >> 1] = (other_bin->data_[idx >> 1] >> ((idx & 1) << 2)) & 0xf;
}
}
void SaveBinaryToFile(const VirtualFileWriter* writer) const override {
writer->Write(data_.data(), sizeof(uint8_t) * data_.size());
}
size_t SizesInByte() const override {
return sizeof(uint8_t) * data_.size();
}
Dense4bitsBin* Clone() override {
return new Dense4bitsBin(*this);
}
protected:
Dense4bitsBin(const Dense4bitsBin& other)
: num_data_(other.num_data_), data_(other.data_), buf_(other.buf_) {}
data_size_t num_data_;
std::vector<uint8_t> data_;
std::vector<uint8_t> buf_;
};
uint32_t Dense4bitsBinIterator::Get(data_size_t idx) {
const auto bin = (bin_data_->data_[idx >> 1] >> ((idx & 1) << 2)) & 0xf;
if (bin >= min_bin_ && bin <= max_bin_) {
return bin - min_bin_ + bias_;
} else {
return default_bin_;
}
}
uint32_t Dense4bitsBinIterator::RawGet(data_size_t idx) {
return (bin_data_->data_[idx >> 1] >> ((idx & 1) << 2)) & 0xf;
}
inline BinIterator* Dense4bitsBin::GetIterator(uint32_t min_bin, uint32_t max_bin, uint32_t default_bin) const {
return new Dense4bitsBinIterator(this, min_bin, max_bin, default_bin);
}
} // namespace LightGBM
#endif // LIGHTGBM_IO_DENSE_NBITS_BIN_HPP_