isotree 0.2.0 → 0.3.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/CHANGELOG.md +16 -1
- data/LICENSE.txt +2 -2
- data/README.md +41 -23
- data/ext/isotree/ext.cpp +144 -31
- data/ext/isotree/extconf.rb +7 -7
- data/lib/isotree/dataset.rb +0 -1
- data/lib/isotree/isolation_forest.rb +114 -30
- data/lib/isotree/version.rb +1 -1
- data/vendor/isotree/LICENSE +1 -1
- data/vendor/isotree/README.md +165 -27
- data/vendor/isotree/include/isotree.hpp +2111 -0
- data/vendor/isotree/include/isotree_oop.hpp +394 -0
- data/vendor/isotree/inst/COPYRIGHTS +62 -0
- data/vendor/isotree/src/RcppExports.cpp +525 -52
- data/vendor/isotree/src/Rwrapper.cpp +1931 -268
- data/vendor/isotree/src/c_interface.cpp +953 -0
- data/vendor/isotree/src/crit.hpp +4232 -0
- data/vendor/isotree/src/dist.hpp +1886 -0
- data/vendor/isotree/src/exp_depth_table.hpp +134 -0
- data/vendor/isotree/src/extended.hpp +1444 -0
- data/vendor/isotree/src/external_facing_generic.hpp +399 -0
- data/vendor/isotree/src/fit_model.hpp +2401 -0
- data/vendor/isotree/src/{dealloc.cpp → headers_joined.hpp} +38 -22
- data/vendor/isotree/src/helpers_iforest.hpp +813 -0
- data/vendor/isotree/src/{impute.cpp → impute.hpp} +353 -122
- data/vendor/isotree/src/indexer.cpp +515 -0
- data/vendor/isotree/src/instantiate_template_headers.cpp +118 -0
- data/vendor/isotree/src/instantiate_template_headers.hpp +240 -0
- data/vendor/isotree/src/isoforest.hpp +1659 -0
- data/vendor/isotree/src/isotree.hpp +1804 -392
- data/vendor/isotree/src/isotree_exportable.hpp +99 -0
- data/vendor/isotree/src/merge_models.cpp +159 -16
- data/vendor/isotree/src/mult.hpp +1321 -0
- data/vendor/isotree/src/oop_interface.cpp +842 -0
- data/vendor/isotree/src/oop_interface.hpp +278 -0
- data/vendor/isotree/src/other_helpers.hpp +219 -0
- data/vendor/isotree/src/predict.hpp +1932 -0
- data/vendor/isotree/src/python_helpers.hpp +134 -0
- data/vendor/isotree/src/ref_indexer.hpp +154 -0
- data/vendor/isotree/src/robinmap/LICENSE +21 -0
- data/vendor/isotree/src/robinmap/README.md +483 -0
- data/vendor/isotree/src/robinmap/include/tsl/robin_growth_policy.h +406 -0
- data/vendor/isotree/src/robinmap/include/tsl/robin_hash.h +1620 -0
- data/vendor/isotree/src/robinmap/include/tsl/robin_map.h +807 -0
- data/vendor/isotree/src/robinmap/include/tsl/robin_set.h +660 -0
- data/vendor/isotree/src/serialize.cpp +4300 -139
- data/vendor/isotree/src/sql.cpp +141 -59
- data/vendor/isotree/src/subset_models.cpp +174 -0
- data/vendor/isotree/src/utils.hpp +3808 -0
- data/vendor/isotree/src/xoshiro.hpp +467 -0
- data/vendor/isotree/src/ziggurat.hpp +405 -0
- metadata +40 -106
- data/vendor/cereal/LICENSE +0 -24
- data/vendor/cereal/README.md +0 -85
- data/vendor/cereal/include/cereal/access.hpp +0 -351
- data/vendor/cereal/include/cereal/archives/adapters.hpp +0 -163
- data/vendor/cereal/include/cereal/archives/binary.hpp +0 -169
- data/vendor/cereal/include/cereal/archives/json.hpp +0 -1019
- data/vendor/cereal/include/cereal/archives/portable_binary.hpp +0 -334
- data/vendor/cereal/include/cereal/archives/xml.hpp +0 -956
- data/vendor/cereal/include/cereal/cereal.hpp +0 -1089
- data/vendor/cereal/include/cereal/details/helpers.hpp +0 -422
- data/vendor/cereal/include/cereal/details/polymorphic_impl.hpp +0 -796
- data/vendor/cereal/include/cereal/details/polymorphic_impl_fwd.hpp +0 -65
- data/vendor/cereal/include/cereal/details/static_object.hpp +0 -127
- data/vendor/cereal/include/cereal/details/traits.hpp +0 -1411
- data/vendor/cereal/include/cereal/details/util.hpp +0 -84
- data/vendor/cereal/include/cereal/external/base64.hpp +0 -134
- data/vendor/cereal/include/cereal/external/rapidjson/allocators.h +0 -284
- data/vendor/cereal/include/cereal/external/rapidjson/cursorstreamwrapper.h +0 -78
- data/vendor/cereal/include/cereal/external/rapidjson/document.h +0 -2652
- data/vendor/cereal/include/cereal/external/rapidjson/encodedstream.h +0 -299
- data/vendor/cereal/include/cereal/external/rapidjson/encodings.h +0 -716
- data/vendor/cereal/include/cereal/external/rapidjson/error/en.h +0 -74
- data/vendor/cereal/include/cereal/external/rapidjson/error/error.h +0 -161
- data/vendor/cereal/include/cereal/external/rapidjson/filereadstream.h +0 -99
- data/vendor/cereal/include/cereal/external/rapidjson/filewritestream.h +0 -104
- data/vendor/cereal/include/cereal/external/rapidjson/fwd.h +0 -151
- data/vendor/cereal/include/cereal/external/rapidjson/internal/biginteger.h +0 -290
- data/vendor/cereal/include/cereal/external/rapidjson/internal/diyfp.h +0 -271
- data/vendor/cereal/include/cereal/external/rapidjson/internal/dtoa.h +0 -245
- data/vendor/cereal/include/cereal/external/rapidjson/internal/ieee754.h +0 -78
- data/vendor/cereal/include/cereal/external/rapidjson/internal/itoa.h +0 -308
- data/vendor/cereal/include/cereal/external/rapidjson/internal/meta.h +0 -186
- data/vendor/cereal/include/cereal/external/rapidjson/internal/pow10.h +0 -55
- data/vendor/cereal/include/cereal/external/rapidjson/internal/regex.h +0 -740
- data/vendor/cereal/include/cereal/external/rapidjson/internal/stack.h +0 -232
- data/vendor/cereal/include/cereal/external/rapidjson/internal/strfunc.h +0 -69
- data/vendor/cereal/include/cereal/external/rapidjson/internal/strtod.h +0 -290
- data/vendor/cereal/include/cereal/external/rapidjson/internal/swap.h +0 -46
- data/vendor/cereal/include/cereal/external/rapidjson/istreamwrapper.h +0 -128
- data/vendor/cereal/include/cereal/external/rapidjson/memorybuffer.h +0 -70
- data/vendor/cereal/include/cereal/external/rapidjson/memorystream.h +0 -71
- data/vendor/cereal/include/cereal/external/rapidjson/msinttypes/inttypes.h +0 -316
- data/vendor/cereal/include/cereal/external/rapidjson/msinttypes/stdint.h +0 -300
- data/vendor/cereal/include/cereal/external/rapidjson/ostreamwrapper.h +0 -81
- data/vendor/cereal/include/cereal/external/rapidjson/pointer.h +0 -1414
- data/vendor/cereal/include/cereal/external/rapidjson/prettywriter.h +0 -277
- data/vendor/cereal/include/cereal/external/rapidjson/rapidjson.h +0 -656
- data/vendor/cereal/include/cereal/external/rapidjson/reader.h +0 -2230
- data/vendor/cereal/include/cereal/external/rapidjson/schema.h +0 -2497
- data/vendor/cereal/include/cereal/external/rapidjson/stream.h +0 -223
- data/vendor/cereal/include/cereal/external/rapidjson/stringbuffer.h +0 -121
- data/vendor/cereal/include/cereal/external/rapidjson/writer.h +0 -709
- data/vendor/cereal/include/cereal/external/rapidxml/license.txt +0 -52
- data/vendor/cereal/include/cereal/external/rapidxml/manual.html +0 -406
- data/vendor/cereal/include/cereal/external/rapidxml/rapidxml.hpp +0 -2624
- data/vendor/cereal/include/cereal/external/rapidxml/rapidxml_iterators.hpp +0 -175
- data/vendor/cereal/include/cereal/external/rapidxml/rapidxml_print.hpp +0 -428
- data/vendor/cereal/include/cereal/external/rapidxml/rapidxml_utils.hpp +0 -123
- data/vendor/cereal/include/cereal/macros.hpp +0 -154
- data/vendor/cereal/include/cereal/specialize.hpp +0 -139
- data/vendor/cereal/include/cereal/types/array.hpp +0 -79
- data/vendor/cereal/include/cereal/types/atomic.hpp +0 -55
- data/vendor/cereal/include/cereal/types/base_class.hpp +0 -203
- data/vendor/cereal/include/cereal/types/bitset.hpp +0 -176
- data/vendor/cereal/include/cereal/types/boost_variant.hpp +0 -164
- data/vendor/cereal/include/cereal/types/chrono.hpp +0 -72
- data/vendor/cereal/include/cereal/types/common.hpp +0 -129
- data/vendor/cereal/include/cereal/types/complex.hpp +0 -56
- data/vendor/cereal/include/cereal/types/concepts/pair_associative_container.hpp +0 -73
- data/vendor/cereal/include/cereal/types/deque.hpp +0 -62
- data/vendor/cereal/include/cereal/types/forward_list.hpp +0 -68
- data/vendor/cereal/include/cereal/types/functional.hpp +0 -43
- data/vendor/cereal/include/cereal/types/list.hpp +0 -62
- data/vendor/cereal/include/cereal/types/map.hpp +0 -36
- data/vendor/cereal/include/cereal/types/memory.hpp +0 -425
- data/vendor/cereal/include/cereal/types/optional.hpp +0 -66
- data/vendor/cereal/include/cereal/types/polymorphic.hpp +0 -483
- data/vendor/cereal/include/cereal/types/queue.hpp +0 -132
- data/vendor/cereal/include/cereal/types/set.hpp +0 -103
- data/vendor/cereal/include/cereal/types/stack.hpp +0 -76
- data/vendor/cereal/include/cereal/types/string.hpp +0 -61
- data/vendor/cereal/include/cereal/types/tuple.hpp +0 -123
- data/vendor/cereal/include/cereal/types/unordered_map.hpp +0 -36
- data/vendor/cereal/include/cereal/types/unordered_set.hpp +0 -99
- data/vendor/cereal/include/cereal/types/utility.hpp +0 -47
- data/vendor/cereal/include/cereal/types/valarray.hpp +0 -89
- data/vendor/cereal/include/cereal/types/variant.hpp +0 -109
- data/vendor/cereal/include/cereal/types/vector.hpp +0 -112
- data/vendor/cereal/include/cereal/version.hpp +0 -52
- data/vendor/isotree/src/Makevars +0 -4
- data/vendor/isotree/src/crit.cpp +0 -912
- data/vendor/isotree/src/dist.cpp +0 -749
- data/vendor/isotree/src/extended.cpp +0 -790
- data/vendor/isotree/src/fit_model.cpp +0 -1090
- data/vendor/isotree/src/helpers_iforest.cpp +0 -324
- data/vendor/isotree/src/isoforest.cpp +0 -771
- data/vendor/isotree/src/mult.cpp +0 -607
- data/vendor/isotree/src/predict.cpp +0 -853
- data/vendor/isotree/src/utils.cpp +0 -1566
@@ -18,11 +18,29 @@
|
|
18
18
|
* [5] https://sourceforge.net/projects/iforest/
|
19
19
|
* [6] https://math.stackexchange.com/questions/3388518/expected-number-of-paths-required-to-separate-elements-in-a-binary-tree
|
20
20
|
* [7] Quinlan, J. Ross. C4. 5: programs for machine learning. Elsevier, 2014.
|
21
|
-
* [8] Cortes, David.
|
22
|
-
*
|
21
|
+
* [8] Cortes, David.
|
22
|
+
* "Distance approximation using Isolation Forests."
|
23
|
+
* arXiv preprint arXiv:1910.12362 (2019).
|
24
|
+
* [9] Cortes, David.
|
25
|
+
* "Imputing missing values with unsupervised random trees."
|
26
|
+
* arXiv preprint arXiv:1911.06646 (2019).
|
27
|
+
* [10] https://math.stackexchange.com/questions/3333220/expected-average-depth-in-random-binary-tree-constructed-top-to-bottom
|
28
|
+
* [11] Cortes, David.
|
29
|
+
* "Revisiting randomized choices in isolation forests."
|
30
|
+
* arXiv preprint arXiv:2110.13402 (2021).
|
31
|
+
* [12] Guha, Sudipto, et al.
|
32
|
+
* "Robust random cut forest based anomaly detection on streams."
|
33
|
+
* International conference on machine learning. PMLR, 2016.
|
34
|
+
* [13] Cortes, David.
|
35
|
+
* "Isolation forests: looking beyond tree depth."
|
36
|
+
* arXiv preprint arXiv:2111.11639 (2021).
|
37
|
+
* [14] Ting, Kai Ming, Yue Zhu, and Zhi-Hua Zhou.
|
38
|
+
* "Isolation kernel and its effect on SVM"
|
39
|
+
* Proceedings of the 24th ACM SIGKDD
|
40
|
+
* International Conference on Knowledge Discovery & Data Mining. 2018.
|
23
41
|
*
|
24
42
|
* BSD 2-Clause License
|
25
|
-
* Copyright (c)
|
43
|
+
* Copyright (c) 2019-2022, David Cortes
|
26
44
|
* All rights reserved.
|
27
45
|
* Redistribution and use in source and binary forms, with or without
|
28
46
|
* modification, are permitted provided that the following conditions are met:
|
@@ -43,73 +61,196 @@
|
|
43
61
|
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
44
62
|
*/
|
45
63
|
|
64
|
+
#ifndef ISOTREE_H
|
65
|
+
#define ISOTREE_H
|
66
|
+
|
67
|
+
/* This is only used for the serialiation format and might not reflect the
|
68
|
+
actual version of the library, do not use for anything else. */
|
69
|
+
#define ISOTREE_VERSION_MAJOR 0
|
70
|
+
#define ISOTREE_VERSION_MINOR 5
|
71
|
+
#define ISOTREE_VERSION_PATCH 6
|
72
|
+
|
73
|
+
/* For MinGW, needs to be defined before including any headers */
|
74
|
+
#if (defined(_WIN32) || defined(_WIN64)) && (SIZE_MAX >= UINT64_MAX)
|
75
|
+
# if defined(__GNUG__) || defined(__GNUC__)
|
76
|
+
# ifndef _FILE_OFFSET_BITS
|
77
|
+
# define _FILE_OFFSET_BITS 64
|
78
|
+
# endif
|
79
|
+
# endif
|
80
|
+
#endif
|
81
|
+
#ifdef _MSC_VER
|
82
|
+
# define _CRT_SECURE_NO_WARNINGS
|
83
|
+
#endif
|
84
|
+
|
85
|
+
|
46
86
|
/* Standard headers */
|
47
|
-
#include <
|
48
|
-
#include <
|
49
|
-
#include <
|
50
|
-
#include <
|
51
|
-
#include <
|
87
|
+
#include <cstddef>
|
88
|
+
#include <cmath>
|
89
|
+
#include <climits>
|
90
|
+
#include <cstring>
|
91
|
+
#include <cerrno>
|
52
92
|
#include <vector>
|
53
93
|
#include <iterator>
|
54
94
|
#include <numeric>
|
55
95
|
#include <algorithm>
|
56
96
|
#include <random>
|
57
|
-
#include <unordered_set>
|
58
|
-
#include <unordered_map>
|
59
97
|
#include <memory>
|
60
98
|
#include <utility>
|
61
99
|
#include <cstdint>
|
100
|
+
#include <cinttypes>
|
101
|
+
#include <exception>
|
102
|
+
#include <stdexcept>
|
103
|
+
#include <cassert>
|
104
|
+
#include <cfloat>
|
62
105
|
#include <iostream>
|
63
|
-
#
|
64
|
-
|
65
|
-
#
|
106
|
+
#include <string>
|
107
|
+
|
108
|
+
#ifdef _FOR_R
|
66
109
|
extern "C" {
|
67
110
|
#include <R_ext/Print.h>
|
68
111
|
}
|
69
112
|
#define printf Rprintf
|
70
113
|
#define fprintf(f, message) REprintf(message)
|
114
|
+
#elif defined(_FOR_PYTHON)
|
115
|
+
extern "C" void cy_warning(const char *msg);
|
116
|
+
#define fprintf(f, message) cy_warning(message)
|
117
|
+
#else
|
118
|
+
#include <cstdio>
|
119
|
+
using std::printf;
|
120
|
+
using std::fprintf;
|
71
121
|
#endif
|
72
122
|
#ifdef _OPENMP
|
73
123
|
#include <omp.h>
|
74
124
|
#endif
|
75
|
-
#ifdef
|
76
|
-
#include <
|
77
|
-
|
78
|
-
|
79
|
-
|
80
|
-
|
125
|
+
#ifdef _FOR_R
|
126
|
+
#include <Rcpp.h>
|
127
|
+
#endif
|
128
|
+
#include <csignal>
|
129
|
+
typedef void (*sig_t_)(int);
|
130
|
+
using std::signal;
|
131
|
+
using std::raise;
|
132
|
+
|
133
|
+
using std::size_t;
|
134
|
+
using std::memset;
|
135
|
+
using std::memcpy;
|
136
|
+
|
137
|
+
#if defined(__GNUC__) || defined(__clang__)
|
138
|
+
#define likely(x) __builtin_expect((bool)(x), true)
|
139
|
+
#define unlikely(x) __builtin_expect((bool)(x), false)
|
140
|
+
#else
|
141
|
+
#define likely(x) (x)
|
142
|
+
#define unlikely(x) (x)
|
143
|
+
#endif
|
144
|
+
|
145
|
+
#if defined(__GNUC__) || defined(__clang__) || defined(_MSC_VER)
|
146
|
+
#define unexpected_error() throw std::runtime_error(\
|
147
|
+
std::string("Unexpected error in ") + \
|
148
|
+
std::string(__FILE__) + \
|
149
|
+
std::string(":") + \
|
150
|
+
std::to_string(__LINE__) + \
|
151
|
+
std::string(". Please open an issue in GitHub with this information, indicating the installed version of 'isotree'.\n"))
|
152
|
+
#else
|
153
|
+
#define unexpected_error() throw std::runtime_error("Unexpected error. Please open an issue in GitHub.\n")
|
81
154
|
#endif
|
82
155
|
|
83
|
-
/* By default, will use
|
84
|
-
#ifdef
|
156
|
+
/* By default, will use Xoshiro256++ or Xoshiro128++ for RNG, but can be switched to something faster */
|
157
|
+
#ifdef _USE_XOSHIRO
|
158
|
+
#include "xoshiro.hpp"
|
85
159
|
#if SIZE_MAX >= UINT64_MAX /* 64-bit systems or higher */
|
86
|
-
#define RNG_engine
|
160
|
+
#define RNG_engine Xoshiro::Xoshiro256PP
|
87
161
|
#else /* 32-bit systems and non-standard architectures */
|
88
|
-
#define RNG_engine
|
162
|
+
#define RNG_engine Xoshiro::Xoshiro128PP
|
163
|
+
#endif
|
164
|
+
#if defined(DBL_MANT_DIG) && (DBL_MANT_DIG == 53) && (FLT_RADIX == 2)
|
165
|
+
using Xoshiro::UniformUnitInterval;
|
166
|
+
using Xoshiro::UniformMinusOneToOne;
|
167
|
+
using Xoshiro::StandardNormalDistr;
|
168
|
+
#else
|
169
|
+
#define UniformUnitInterval std::uniform_real_distribution<double>
|
170
|
+
#define UniformMinusOneToOne std::uniform_real_distribution<double>
|
171
|
+
#define StandardNormalDistr std::normal_distribution<double>
|
172
|
+
#endif
|
173
|
+
#else
|
174
|
+
#if defined(_USE_MERSENNE_TWISTER)
|
175
|
+
#if SIZE_MAX >= UINT64_MAX /* 64-bit systems or higher */
|
176
|
+
#define RNG_engine std::mt19937_64
|
177
|
+
#else /* 32-bit systems and non-standard architectures */
|
178
|
+
#define RNG_engine std::mt19937
|
179
|
+
#endif
|
180
|
+
#else
|
181
|
+
#define RNG_engine std::default_random_engine
|
182
|
+
#endif
|
183
|
+
|
184
|
+
#define UniformUnitInterval std::uniform_real_distribution<double>
|
185
|
+
#define UniformMinusOneToOne std::uniform_real_distribution<double>
|
186
|
+
#define StandardNormalDistr std::normal_distribution<double>
|
187
|
+
#endif
|
188
|
+
|
189
|
+
/* At the time of writing, this brought a sizeable speed up compared to
|
190
|
+
'unordered_map' and 'unordered_set' from both GCC and CLANG.
|
191
|
+
But perhaps should consider others in the future, such as this:
|
192
|
+
https://github.com/ktprime/emhash */
|
193
|
+
#if defined(_USE_ROBIN_MAP)
|
194
|
+
#ifndef _USE_SYSTEM_ROBIN
|
195
|
+
#include "robinmap/include/tsl/robin_growth_policy.h"
|
196
|
+
#include "robinmap/include/tsl/robin_hash.h"
|
197
|
+
#include "robinmap/include/tsl/robin_set.h"
|
198
|
+
#include "robinmap/include/tsl/robin_map.h"
|
199
|
+
#else
|
200
|
+
#include "tsl/robin_growth_policy.h"
|
201
|
+
#include "tsl/robin_hash.h"
|
202
|
+
#include "tsl/robin_set.h"
|
203
|
+
#include "tsl/robin_map.h"
|
89
204
|
#endif
|
205
|
+
#define hashed_set tsl::robin_set
|
206
|
+
#define hashed_map tsl::robin_map
|
90
207
|
#else
|
91
|
-
#
|
208
|
+
#include <unordered_set>
|
209
|
+
#include <unordered_map>
|
210
|
+
#define hashed_set std::unordered_set
|
211
|
+
#define hashed_map std::unordered_map
|
92
212
|
#endif
|
93
213
|
|
94
214
|
/* Short functions */
|
95
|
-
#define ix_parent(ix) (((ix) - 1) / 2) /* integer division takes care of deciding left-right */
|
96
|
-
#define ix_child(ix) (2 * (ix) + 1)
|
97
215
|
/* https://stackoverflow.com/questions/101439/the-most-efficient-way-to-implement-an-integer-based-power-function-powint-int */
|
98
216
|
#define pow2(n) ( ((size_t) 1) << (n) )
|
217
|
+
#define div2(n) ((n) >> 1)
|
218
|
+
#define mult2(n) ((n) << 1)
|
219
|
+
#define ix_parent(ix) (div2((ix) - (size_t)1)) /* integer division takes care of deciding left-right */
|
220
|
+
#define ix_child(ix) (mult2(ix) + (size_t)1)
|
99
221
|
#define square(x) ((x) * (x))
|
222
|
+
#ifndef _FOR_R
|
223
|
+
#if defined(__GNUC__) && (__GNUC__ >= 5)
|
224
|
+
#pragma GCC diagnostic push
|
225
|
+
#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
|
226
|
+
#elif defined(__clang__) && !defined(_FOR_R)
|
227
|
+
#pragma clang diagnostic push
|
228
|
+
#pragma clang diagnostic ignored "-Wuninitialized"
|
229
|
+
#endif
|
230
|
+
#endif
|
100
231
|
/* https://stackoverflow.com/questions/2249731/how-do-i-get-bit-by-bit-data-from-an-integer-value-in-c */
|
101
232
|
#define extract_bit(number, bit) (((number) >> (bit)) & 1)
|
102
|
-
#ifndef
|
103
|
-
#
|
233
|
+
#ifndef _FOR_R
|
234
|
+
#if defined(__GNUC__) && (__GNUC__ >= 5)
|
235
|
+
#pragma GCC diagnostic pop
|
236
|
+
#elif defined(__clang__)
|
237
|
+
#pragma clang diagnostic pop
|
238
|
+
#pragma clang diagnostic push
|
239
|
+
#pragma clang diagnostic ignored "-Wunknown-attributes"
|
240
|
+
#endif
|
104
241
|
#endif
|
105
|
-
#
|
106
|
-
|
242
|
+
#define is_na_or_inf(x) (std::isnan(x) || std::isinf(x))
|
243
|
+
|
244
|
+
/* MSVC doesn't support long doubles, so this avoids unnecessarily increasing library size.
|
245
|
+
MinGW supports them but has issues with their computations.
|
246
|
+
See https://sourceforge.net/p/mingw-w64/bugs/909/ */
|
247
|
+
#if defined(_WIN32) && !defined(NO_LONG_DOUBLE)
|
248
|
+
#define NO_LONG_DOUBLE
|
107
249
|
#endif
|
108
|
-
#define is_na_or_inf(x) (isnan(x) || isinf(x))
|
109
250
|
|
110
251
|
|
111
252
|
/* Aliasing for compiler optimizations */
|
112
|
-
#if defined(__GNUG__) || defined(__GNUC__) || defined(_MSC_VER) || defined(__clang__) || defined(__INTEL_COMPILER)
|
253
|
+
#if defined(__GNUG__) || defined(__GNUC__) || defined(_MSC_VER) || defined(__clang__) || defined(__INTEL_COMPILER) || defined(__IBMCPP__) || defined(__ibmxl__) || defined(SUPPORTS_RESTRICT)
|
113
254
|
#define restrict __restrict
|
114
255
|
#else
|
115
256
|
#define restrict
|
@@ -118,7 +259,7 @@
|
|
118
259
|
/* MSVC is stuck with an OpenMP version that's 19 years old at the time of writing and does not support unsigned iterators */
|
119
260
|
#ifdef _OPENMP
|
120
261
|
#if (_OPENMP < 200801) || defined(_WIN32) || defined(_WIN64) /* OpenMP < 3.0 */
|
121
|
-
#define size_t_for long
|
262
|
+
#define size_t_for long long
|
122
263
|
#else
|
123
264
|
#define size_t_for size_t
|
124
265
|
#endif
|
@@ -126,33 +267,51 @@
|
|
126
267
|
#define size_t_for size_t
|
127
268
|
#endif
|
128
269
|
|
270
|
+
#if defined(_FOR_R) || defined(_FOR_PYTHON)
|
271
|
+
#define ISOTREE_EXPORTED
|
272
|
+
#else
|
273
|
+
#if defined(_WIN32)
|
274
|
+
#ifdef ISOTREE_COMPILE_TIME
|
275
|
+
#define ISOTREE_EXPORTED __declspec(dllexport)
|
276
|
+
#else
|
277
|
+
#define ISOTREE_EXPORTED __declspec(dllimport)
|
278
|
+
#endif
|
279
|
+
#else
|
280
|
+
#if defined(EXPLICITLTY_EXPORT_SYMBOLS) && defined(ISOTREE_COMPILE_TIME)
|
281
|
+
#define ISOTREE_EXPORTED [[gnu::visibility("default")]]
|
282
|
+
#else
|
283
|
+
#define ISOTREE_EXPORTED
|
284
|
+
#endif
|
285
|
+
#endif
|
286
|
+
#endif
|
287
|
+
|
129
288
|
|
130
|
-
/*
|
289
|
+
/* Apple at some point decided to drop OMP library and headers from its compiler distribution
|
131
290
|
* and to alias 'gcc' to 'clang', which work differently when given flags they cannot interpret,
|
132
291
|
* causing installation issues with pretty much all scientific software due to OMP headers that
|
133
292
|
* would normally do nothing. This piece of code is to allow compilation without OMP header. */
|
134
293
|
#ifndef _OPENMP
|
135
|
-
#define omp_get_thread_num() 0
|
294
|
+
#define omp_get_thread_num() (0)
|
136
295
|
#endif
|
137
296
|
|
297
|
+
/* Some aggregation functions will prefer more precise data types when the data is large */
|
298
|
+
#define THRESHOLD_LONG_DOUBLE (size_t)1e6
|
138
299
|
|
139
|
-
/*
|
140
|
-
|
141
|
-
|
142
|
-
|
143
|
-
|
144
|
-
|
300
|
+
/* Types used through the package */
|
301
|
+
typedef enum NewCategAction {Weighted=0, Smallest=11, Random=12} NewCategAction; /* Weighted means Impute in the extended model */
|
302
|
+
typedef enum MissingAction {Divide=21, Impute=22, Fail=0} MissingAction; /* Divide is only for non-extended model */
|
303
|
+
typedef enum ColType {Numeric=31, Categorical=32, NotUsed=0} ColType;
|
304
|
+
typedef enum CategSplit {SubSet=0, SingleCateg=41} CategSplit;
|
305
|
+
typedef enum CoefType {Uniform=61, Normal=0} CoefType; /* For extended model */
|
306
|
+
typedef enum UseDepthImp {Lower=71, Higher=0, Same=72} UseDepthImp; /* For NA imputation */
|
307
|
+
typedef enum WeighImpRows {Inverse=0, Prop=81, Flat=82} WeighImpRows; /* For NA imputation */
|
308
|
+
typedef enum ScoringMetric {Depth=0, Density=92, BoxedDensity=94, BoxedDensity2=96, BoxedRatio=95,
|
309
|
+
AdjDepth=91, AdjDensity=93} ScoringMetric;
|
145
310
|
|
311
|
+
/* These are only used internally */
|
312
|
+
typedef enum ColCriterion {Uniformly=0, ByRange=1, ByVar=2, ByKurt=3} ColCriterion; /* For proportional choices */
|
313
|
+
typedef enum GainCriterion {NoCrit=0, Averaged=1, Pooled=2, FullGain=3, DensityCrit=4} Criterion; /* For guided splits */
|
146
314
|
|
147
|
-
/* Types used through the package */
|
148
|
-
typedef enum NewCategAction {Weighted, Smallest, Random} NewCategAction; /* Weighted means Impute in the extended model */
|
149
|
-
typedef enum MissingAction {Divide, Impute, Fail} MissingAction; /* Divide is only for non-extended model */
|
150
|
-
typedef enum ColType {Numeric, Categorical, NotUsed} ColType;
|
151
|
-
typedef enum CategSplit {SubSet, SingleCateg} CategSplit;
|
152
|
-
typedef enum GainCriterion {Averaged, Pooled, NoCrit} Criterion; /* For guided splits */
|
153
|
-
typedef enum CoefType {Uniform, Normal} CoefType; /* For extended model */
|
154
|
-
typedef enum UseDepthImp {Lower, Higher, Same} UseDepthImp; /* For NA imputation */
|
155
|
-
typedef enum WeighImpRows {Inverse, Prop, Flat} WeighImpRows; /* For NA imputation */
|
156
315
|
|
157
316
|
/* Notes about new categorical action:
|
158
317
|
* - For single-variable case, if using 'Smallest', can then pass data at prediction time
|
@@ -167,10 +326,10 @@ typedef enum WeighImpRows {Inverse, Prop, Flat} WeighImpRows; /
|
|
167
326
|
|
168
327
|
/* Structs that are output (modified) from the main function */
|
169
328
|
typedef struct IsoTree {
|
170
|
-
ColType col_type = NotUsed; /* issues with uninitialized values
|
329
|
+
ColType col_type = NotUsed; /* issues with uninitialized values when serializing */
|
171
330
|
size_t col_num;
|
172
331
|
double num_split;
|
173
|
-
std::vector<char> cat_split;
|
332
|
+
std::vector<signed char> cat_split;
|
174
333
|
int chosen_cat;
|
175
334
|
size_t tree_left;
|
176
335
|
size_t tree_right;
|
@@ -180,29 +339,7 @@ typedef struct IsoTree {
|
|
180
339
|
double range_high = HUGE_VAL;
|
181
340
|
double remainder; /* only used for distance/similarity */
|
182
341
|
|
183
|
-
#ifdef _ENABLE_CEREAL
|
184
|
-
template<class Archive>
|
185
|
-
void serialize(Archive &archive)
|
186
|
-
{
|
187
|
-
archive(
|
188
|
-
this->col_type,
|
189
|
-
this->col_num,
|
190
|
-
this->num_split,
|
191
|
-
this->cat_split,
|
192
|
-
this->chosen_cat,
|
193
|
-
this->tree_left,
|
194
|
-
this->tree_right,
|
195
|
-
this->pct_tree_left,
|
196
|
-
this->score,
|
197
|
-
this->range_low,
|
198
|
-
this->range_high,
|
199
|
-
this->remainder
|
200
|
-
);
|
201
|
-
}
|
202
|
-
#endif
|
203
|
-
|
204
342
|
IsoTree() = default;
|
205
|
-
|
206
343
|
} IsoTree;
|
207
344
|
|
208
345
|
typedef struct IsoHPlane {
|
@@ -223,30 +360,6 @@ typedef struct IsoHPlane {
|
|
223
360
|
double range_high = HUGE_VAL;
|
224
361
|
double remainder; /* only used for distance/similarity */
|
225
362
|
|
226
|
-
#ifdef _ENABLE_CEREAL
|
227
|
-
template<class Archive>
|
228
|
-
void serialize(Archive &archive)
|
229
|
-
{
|
230
|
-
archive(
|
231
|
-
this->col_num,
|
232
|
-
this->col_type,
|
233
|
-
this->coef,
|
234
|
-
this->mean,
|
235
|
-
this->cat_coef,
|
236
|
-
this->chosen_cat,
|
237
|
-
this->fill_val,
|
238
|
-
this->fill_new,
|
239
|
-
this->split_point,
|
240
|
-
this->hplane_left,
|
241
|
-
this->hplane_right,
|
242
|
-
this->score,
|
243
|
-
this->range_low,
|
244
|
-
this->range_high,
|
245
|
-
this->remainder
|
246
|
-
);
|
247
|
-
}
|
248
|
-
#endif
|
249
|
-
|
250
363
|
IsoHPlane() = default;
|
251
364
|
} IsoHPlane;
|
252
365
|
|
@@ -258,25 +371,11 @@ typedef struct IsoForest {
|
|
258
371
|
NewCategAction new_cat_action;
|
259
372
|
CategSplit cat_split_type;
|
260
373
|
MissingAction missing_action;
|
374
|
+
ScoringMetric scoring_metric;
|
261
375
|
double exp_avg_depth;
|
262
376
|
double exp_avg_sep;
|
263
377
|
size_t orig_sample_size;
|
264
|
-
|
265
|
-
#ifdef _ENABLE_CEREAL
|
266
|
-
template<class Archive>
|
267
|
-
void serialize(Archive &archive)
|
268
|
-
{
|
269
|
-
archive(
|
270
|
-
this->trees,
|
271
|
-
this->new_cat_action,
|
272
|
-
this->cat_split_type,
|
273
|
-
this->missing_action,
|
274
|
-
this->exp_avg_depth,
|
275
|
-
this->exp_avg_sep,
|
276
|
-
this->orig_sample_size
|
277
|
-
);
|
278
|
-
}
|
279
|
-
#endif
|
378
|
+
bool has_range_penalty;
|
280
379
|
|
281
380
|
IsoForest() = default;
|
282
381
|
} IsoForest;
|
@@ -286,25 +385,11 @@ typedef struct ExtIsoForest {
|
|
286
385
|
NewCategAction new_cat_action;
|
287
386
|
CategSplit cat_split_type;
|
288
387
|
MissingAction missing_action;
|
388
|
+
ScoringMetric scoring_metric;
|
289
389
|
double exp_avg_depth;
|
290
390
|
double exp_avg_sep;
|
291
391
|
size_t orig_sample_size;
|
292
|
-
|
293
|
-
#ifdef _ENABLE_CEREAL
|
294
|
-
template<class Archive>
|
295
|
-
void serialize(Archive &archive)
|
296
|
-
{
|
297
|
-
archive(
|
298
|
-
this->hplanes,
|
299
|
-
this->new_cat_action,
|
300
|
-
this->cat_split_type,
|
301
|
-
this->missing_action,
|
302
|
-
this->exp_avg_depth,
|
303
|
-
this->exp_avg_sep,
|
304
|
-
this->orig_sample_size
|
305
|
-
);
|
306
|
-
}
|
307
|
-
#endif
|
392
|
+
bool has_range_penalty;
|
308
393
|
|
309
394
|
ExtIsoForest() = default;
|
310
395
|
} ExtIsoForest;
|
@@ -316,19 +401,6 @@ typedef struct ImputeNode {
|
|
316
401
|
std::vector<double> cat_weight;
|
317
402
|
size_t parent;
|
318
403
|
|
319
|
-
#ifdef _ENABLE_CEREAL
|
320
|
-
template<class Archive>
|
321
|
-
void serialize(Archive &archive)
|
322
|
-
{
|
323
|
-
archive(
|
324
|
-
this->num_sum,
|
325
|
-
this->num_weight,
|
326
|
-
this->cat_sum,
|
327
|
-
this->cat_weight,
|
328
|
-
this->parent
|
329
|
-
);
|
330
|
-
}
|
331
|
-
#endif
|
332
404
|
ImputeNode() = default;
|
333
405
|
|
334
406
|
ImputeNode(size_t parent)
|
@@ -345,30 +417,31 @@ typedef struct Imputer {
|
|
345
417
|
std::vector<std::vector<ImputeNode>> imputer_tree;
|
346
418
|
std::vector<double> col_means;
|
347
419
|
std::vector<int> col_modes;
|
420
|
+
|
421
|
+
Imputer() = default;
|
422
|
+
} Imputer;
|
348
423
|
|
349
|
-
|
350
|
-
|
351
|
-
|
352
|
-
|
353
|
-
|
354
|
-
|
355
|
-
|
356
|
-
|
357
|
-
|
358
|
-
this->col_means,
|
359
|
-
this->col_modes
|
360
|
-
);
|
361
|
-
}
|
362
|
-
#endif
|
424
|
+
typedef struct SingleTreeIndex {
|
425
|
+
std::vector<size_t> terminal_node_mappings;
|
426
|
+
std::vector<double> node_distances;
|
427
|
+
std::vector<double> node_depths;
|
428
|
+
std::vector<size_t> reference_points;
|
429
|
+
std::vector<size_t> reference_indptr;
|
430
|
+
std::vector<size_t> reference_mapping;
|
431
|
+
size_t n_terminal;
|
432
|
+
} TreeNodeIndex;
|
363
433
|
|
364
|
-
|
434
|
+
typedef struct TreesIndexer {
|
435
|
+
std::vector<SingleTreeIndex> indices;
|
365
436
|
|
366
|
-
|
437
|
+
TreesIndexer() = default;
|
438
|
+
} TreesIndexer;
|
367
439
|
|
368
440
|
|
369
441
|
/* Structs that are only used internally */
|
370
|
-
|
371
|
-
|
442
|
+
template <class real_t, class sparse_ix>
|
443
|
+
struct InputData {
|
444
|
+
real_t* numeric_data;
|
372
445
|
size_t ncols_numeric;
|
373
446
|
int* categ_data;
|
374
447
|
int* ncat;
|
@@ -376,10 +449,10 @@ typedef struct {
|
|
376
449
|
size_t ncols_categ;
|
377
450
|
size_t nrows;
|
378
451
|
size_t ncols_tot;
|
379
|
-
|
452
|
+
real_t* sample_weights;
|
380
453
|
bool weight_as_sample;
|
381
|
-
|
382
|
-
|
454
|
+
real_t* col_weights;
|
455
|
+
real_t* Xc; /* only for sparse matrices */
|
383
456
|
sparse_ix* Xc_ind; /* only for sparse matrices */
|
384
457
|
sparse_ix* Xc_indptr; /* only for sparse matrices */
|
385
458
|
size_t log2_n; /* only when using weights for sampling */
|
@@ -387,37 +460,58 @@ typedef struct {
|
|
387
460
|
std::vector<double> btree_weights_init; /* only when using weights for sampling */
|
388
461
|
std::vector<char> has_missing; /* only used when producing missing imputations on-the-fly */
|
389
462
|
size_t n_missing; /* only used when producing missing imputations on-the-fly */
|
390
|
-
|
391
|
-
|
392
|
-
|
393
|
-
|
394
|
-
double*
|
463
|
+
void* preinitialized_col_sampler; /* only when using column weights */
|
464
|
+
double* range_low; /* only when calculating variable ranges or boxed densities with no sub-sampling */
|
465
|
+
double* range_high; /* only when calculating variable ranges or boxed densities with no sub-sampling */
|
466
|
+
int* ncat_; /* only when calculating boxed densities with no sub-sampling */
|
467
|
+
std::vector<double> all_kurtoses; /* only when using 'prob_pick_col_by_kurtosis' or mixing 'weigh_by_kurt' with 'prob_pick_col*' with no sub-sampling */
|
468
|
+
|
469
|
+
std::vector<double> X_row_major; /* created by this library, only used when calculating full gain */
|
470
|
+
std::vector<double> Xr; /* created by this library, only used when calculating full gain */
|
471
|
+
std::vector<size_t> Xr_ind; /* created by this library, only used when calculating full gain */
|
472
|
+
std::vector<size_t> Xr_indptr; /* created by this library, only used when calculating full gain */
|
473
|
+
};
|
474
|
+
|
475
|
+
|
476
|
+
template <class real_t, class sparse_ix>
|
477
|
+
struct PredictionData {
|
478
|
+
real_t* numeric_data;
|
395
479
|
int* categ_data;
|
396
480
|
size_t nrows;
|
397
|
-
|
398
|
-
|
399
|
-
|
400
|
-
|
401
|
-
sparse_ix*
|
402
|
-
sparse_ix*
|
403
|
-
|
481
|
+
bool is_col_major;
|
482
|
+
size_t ncols_numeric; /* only required for row-major data */
|
483
|
+
size_t ncols_categ; /* only required for row-major data */
|
484
|
+
real_t* Xc; /* only for sparse matrices */
|
485
|
+
sparse_ix* Xc_ind; /* only for sparse matrices */
|
486
|
+
sparse_ix* Xc_indptr; /* only for sparse matrices */
|
487
|
+
real_t* Xr; /* only for sparse matrices */
|
488
|
+
sparse_ix* Xr_ind; /* only for sparse matrices */
|
489
|
+
sparse_ix* Xr_indptr; /* only for sparse matrices */
|
490
|
+
};
|
404
491
|
|
405
492
|
typedef struct {
|
406
493
|
bool with_replacement;
|
407
494
|
size_t sample_size;
|
408
495
|
size_t ntrees;
|
496
|
+
size_t ncols_per_tree;
|
409
497
|
size_t max_depth;
|
410
498
|
bool penalize_range;
|
499
|
+
bool standardize_data;
|
411
500
|
uint64_t random_seed;
|
412
501
|
bool weigh_by_kurt;
|
413
502
|
double prob_pick_by_gain_avg;
|
414
|
-
double prob_split_by_gain_avg;
|
415
503
|
double prob_pick_by_gain_pl;
|
416
|
-
double
|
504
|
+
double prob_pick_by_full_gain;
|
505
|
+
double prob_pick_by_dens;
|
506
|
+
double prob_pick_col_by_range;
|
507
|
+
double prob_pick_col_by_var;
|
508
|
+
double prob_pick_col_by_kurt;
|
417
509
|
double min_gain;
|
418
510
|
CategSplit cat_split_type;
|
419
511
|
NewCategAction new_cat_action;
|
420
512
|
MissingAction missing_action;
|
513
|
+
ScoringMetric scoring_metric;
|
514
|
+
bool fast_bratio;
|
421
515
|
bool all_perm;
|
422
516
|
|
423
517
|
size_t ndim; /* only for extended model */
|
@@ -431,16 +525,17 @@ typedef struct {
|
|
431
525
|
|
432
526
|
UseDepthImp depth_imp; /* only when building NA imputer */
|
433
527
|
WeighImpRows weigh_imp_rows; /* only when building NA imputer */
|
434
|
-
size_t min_imp_obs;
|
528
|
+
size_t min_imp_obs; /* only when building NA imputer */
|
435
529
|
} ModelParams;
|
436
530
|
|
437
|
-
|
438
|
-
|
439
|
-
std::vector<
|
440
|
-
std::vector<
|
441
|
-
std::vector<
|
442
|
-
std::vector<
|
443
|
-
std::vector<
|
531
|
+
template <class sparse_ix, class ldouble_safe>
|
532
|
+
struct ImputedData {
|
533
|
+
std::vector<ldouble_safe> num_sum;
|
534
|
+
std::vector<ldouble_safe> num_weight;
|
535
|
+
std::vector<std::vector<ldouble_safe>> cat_sum;
|
536
|
+
std::vector<ldouble_safe> cat_weight;
|
537
|
+
std::vector<ldouble_safe> sp_num_sum;
|
538
|
+
std::vector<ldouble_safe> sp_num_weight;
|
444
539
|
|
445
540
|
std::vector<size_t> missing_num;
|
446
541
|
std::vector<size_t> missing_cat;
|
@@ -451,56 +546,288 @@ typedef struct ImputedData {
|
|
451
546
|
|
452
547
|
ImputedData() {};
|
453
548
|
|
454
|
-
|
455
|
-
|
456
|
-
|
549
|
+
template <class InputData>
|
550
|
+
ImputedData(InputData &input_data, size_t row)
|
551
|
+
{
|
552
|
+
initialize_impute_calc(*this, input_data, row);
|
553
|
+
}
|
457
554
|
|
458
|
-
|
555
|
+
};
|
556
|
+
|
557
|
+
/* This class provides efficient methods for sampling columns at random,
|
558
|
+
given that at a given node a column might no longer be splittable,
|
559
|
+
and when that happens, it also makes it non-splittable in any children
|
560
|
+
node from there onwards. The idea is to provide efficient methods for
|
561
|
+
passing the state from a parent node to a left node and then restore
|
562
|
+
the state before going for the right node.
|
563
|
+
It can be used in 3 modes:
|
564
|
+
- As a uniform sampler with replacement.
|
565
|
+
- As a weighted sampler with replacement.
|
566
|
+
- As an array that keeps track of which columns are still splittable. */
|
567
|
+
template <class ldouble_safe>
|
568
|
+
class ColumnSampler
|
569
|
+
{
|
570
|
+
public:
|
571
|
+
std::vector<size_t> col_indices;
|
572
|
+
std::vector<double> tree_weights;
|
573
|
+
size_t curr_pos;
|
574
|
+
size_t curr_col;
|
575
|
+
size_t last_given;
|
576
|
+
size_t n_cols;
|
577
|
+
size_t tree_levels;
|
578
|
+
size_t offset;
|
579
|
+
size_t n_dropped;
|
580
|
+
template <class real_t>
|
581
|
+
void initialize(real_t weights[], size_t n_cols);
|
582
|
+
void initialize(size_t n_cols);
|
583
|
+
void drop_weights();
|
584
|
+
void leave_m_cols(size_t m, RNG_engine &rnd_generator);
|
585
|
+
bool sample_col(size_t &col, RNG_engine &rnd_generator);
|
586
|
+
void prepare_full_pass(); /* when passing through all columns */
|
587
|
+
bool sample_col(size_t &col); /* when passing through all columns */
|
588
|
+
void drop_col(size_t col, size_t nobs_left);
|
589
|
+
void drop_col(size_t col);
|
590
|
+
void drop_from_tail(size_t col);
|
591
|
+
void shuffle_remainder(RNG_engine &rnd_generator);
|
592
|
+
bool has_weights();
|
593
|
+
size_t get_remaining_cols();
|
594
|
+
void get_array_remaining_cols(std::vector<size_t> &restrict cols);
|
595
|
+
template <class other_t>
|
596
|
+
ColumnSampler& operator=(const ColumnSampler<other_t> &other);
|
597
|
+
ColumnSampler() = default;
|
598
|
+
};
|
599
|
+
|
600
|
+
template <class ldouble_safe, class real_t>
|
601
|
+
class DensityCalculator
|
602
|
+
{
|
603
|
+
public:
|
604
|
+
std::vector<ldouble_safe> multipliers;
|
605
|
+
double xmin;
|
606
|
+
double xmax;
|
607
|
+
std::vector<size_t> counts;
|
608
|
+
int n_present;
|
609
|
+
int n_left;
|
610
|
+
std::vector<double> box_low;
|
611
|
+
std::vector<double> box_high;
|
612
|
+
std::vector<double> queue_box;
|
613
|
+
bool fast_bratio;
|
614
|
+
std::vector<ldouble_safe> ranges;
|
615
|
+
std::vector<int> ncat;
|
616
|
+
std::vector<int> queue_ncat;
|
617
|
+
std::vector<int> ncat_orig;
|
618
|
+
std::vector<double> vals_ext_box;
|
619
|
+
std::vector<double> queue_ext_box;
|
620
|
+
|
621
|
+
void initialize(size_t max_depth, int max_categ, bool reserve_counts, ScoringMetric scoring_metric);
|
622
|
+
template <class InputData>
|
623
|
+
#ifndef _FOR_R
|
624
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
625
|
+
#endif
|
626
|
+
void initialize_bdens(const InputData &input_data,
|
627
|
+
const ModelParams &model_params,
|
628
|
+
std::vector<size_t> &ix_arr,
|
629
|
+
ColumnSampler<ldouble_safe> &col_sampler);
|
630
|
+
template <class InputData>
|
631
|
+
void initialize_bdens_ext(const InputData &input_data,
|
632
|
+
const ModelParams &model_params,
|
633
|
+
std::vector<size_t> &ix_arr,
|
634
|
+
ColumnSampler<ldouble_safe> &col_sampler,
|
635
|
+
bool col_sampler_is_fresh);
|
636
|
+
#ifndef _FOR_R
|
637
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
638
|
+
#endif
|
639
|
+
void push_density(double xmin, double xmax, double split_point);
|
640
|
+
void push_density(size_t counts[], int ncat);
|
641
|
+
void push_density(int n_left, int n_present);
|
642
|
+
void push_density(int n_present);
|
643
|
+
void push_density();
|
644
|
+
void push_adj(double xmin, double xmax, double split_point, double pct_tree_left, ScoringMetric scoring_metric);
|
645
|
+
void push_adj(signed char *restrict categ_present, size_t *restrict counts, int ncat, ScoringMetric scoring_metric);
|
646
|
+
void push_adj(size_t *restrict counts, int ncat, int chosen_cat, ScoringMetric scoring_metric);
|
647
|
+
void push_adj(double pct_tree_left, ScoringMetric scoring_metric);
|
648
|
+
void push_bdens(double split_point, size_t col);
|
649
|
+
void push_bdens(int ncat_branch_left, size_t col);
|
650
|
+
void push_bdens(const std::vector<signed char> &cat_split, size_t col);
|
651
|
+
#ifndef _FOR_R
|
652
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
653
|
+
#endif
|
654
|
+
void push_bdens_fast_route(double split_point, size_t col);
|
655
|
+
void push_bdens_internal(double split_point, size_t col);
|
656
|
+
#ifndef _FOR_R
|
657
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
658
|
+
#endif
|
659
|
+
void push_bdens_fast_route(int ncat_branch_left, size_t col);
|
660
|
+
void push_bdens_internal(int ncat_branch_left, size_t col);
|
661
|
+
#ifndef _FOR_R
|
662
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
663
|
+
#endif
|
664
|
+
void push_bdens_fast_route(const std::vector<signed char> &cat_split, size_t col);
|
665
|
+
void push_bdens_internal(const std::vector<signed char> &cat_split, size_t col);
|
666
|
+
#ifndef _FOR_R
|
667
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
668
|
+
#endif
|
669
|
+
void push_bdens_ext(const IsoHPlane &hplane, const ModelParams &model_params);
|
670
|
+
void pop();
|
671
|
+
void pop_right();
|
672
|
+
void pop_bdens(size_t col);
|
673
|
+
void pop_bdens_right(size_t col);
|
674
|
+
void pop_bdens_cat(size_t col);
|
675
|
+
void pop_bdens_cat_right(size_t col);
|
676
|
+
void pop_bdens_fast_route(size_t col);
|
677
|
+
void pop_bdens_internal(size_t col);
|
678
|
+
void pop_bdens_right_fast_route(size_t col);
|
679
|
+
void pop_bdens_right_internal(size_t col);
|
680
|
+
void pop_bdens_cat_fast_route(size_t col);
|
681
|
+
void pop_bdens_cat_internal(size_t col);
|
682
|
+
void pop_bdens_cat_right_fast_route(size_t col);
|
683
|
+
void pop_bdens_cat_right_internal(size_t col);
|
684
|
+
void pop_bdens_ext();
|
685
|
+
void pop_bdens_ext_right();
|
686
|
+
#ifndef _FOR_R
|
687
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
688
|
+
#endif
|
689
|
+
double calc_density(ldouble_safe remainder, size_t sample_size);
|
690
|
+
ldouble_safe calc_adj_depth();
|
691
|
+
double calc_adj_density();
|
692
|
+
#ifndef _FOR_R
|
693
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
694
|
+
#endif
|
695
|
+
ldouble_safe calc_bratio_log();
|
696
|
+
#ifndef _FOR_R
|
697
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
698
|
+
#endif
|
699
|
+
ldouble_safe calc_bratio_inv_log();
|
700
|
+
#ifndef _FOR_R
|
701
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
702
|
+
#endif
|
703
|
+
double calc_bratio();
|
704
|
+
#ifndef _FOR_R
|
705
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
706
|
+
#endif
|
707
|
+
double calc_bdens(ldouble_safe remainder, size_t sample_size);
|
708
|
+
#ifndef _FOR_R
|
709
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
710
|
+
#endif
|
711
|
+
double calc_bdens2(ldouble_safe remainder, size_t sample_size);
|
712
|
+
#ifndef _FOR_R
|
713
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
714
|
+
#endif
|
715
|
+
ldouble_safe calc_bratio_log_ext();
|
716
|
+
#ifndef _FOR_R
|
717
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
718
|
+
#endif
|
719
|
+
double calc_bratio_ext();
|
720
|
+
#ifndef _FOR_R
|
721
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
722
|
+
#endif
|
723
|
+
double calc_bdens_ext(ldouble_safe remainder, size_t sample_size);
|
724
|
+
void save_range(double xmin, double xmax);
|
725
|
+
void restore_range(double &restrict xmin, double &restrict xmax);
|
726
|
+
void save_counts(size_t *restrict cat_counts, int ncat);
|
727
|
+
void save_n_present_and_left(signed char *restrict split_left, int ncat);
|
728
|
+
void save_n_present(size_t *restrict cat_counts, int ncat);
|
729
|
+
};
|
730
|
+
|
731
|
+
template <class ldouble_safe, class real_t>
|
732
|
+
class SingleNodeColumnSampler
|
733
|
+
{
|
734
|
+
public:
|
735
|
+
double *restrict weights_orig;
|
736
|
+
std::vector<bool> inifinite_weights;
|
737
|
+
ldouble_safe cumw;
|
738
|
+
size_t n_inf;
|
739
|
+
size_t *restrict col_indices;
|
740
|
+
size_t curr_pos;
|
741
|
+
bool using_tree;
|
742
|
+
|
743
|
+
bool backup_weights;
|
744
|
+
std::vector<double> weights_own;
|
745
|
+
size_t n_left;
|
746
|
+
|
747
|
+
std::vector<double> tree_weights;
|
748
|
+
size_t offset;
|
749
|
+
size_t tree_levels;
|
750
|
+
std::vector<double> used_weights;
|
751
|
+
std::vector<size_t> mapped_indices;
|
752
|
+
std::vector<size_t> mapped_inf_indices;
|
753
|
+
|
754
|
+
bool initialize(
|
755
|
+
double *restrict weights,
|
756
|
+
std::vector<size_t> *col_indices,
|
757
|
+
size_t curr_pos,
|
758
|
+
size_t n_sample,
|
759
|
+
bool backup_weights
|
760
|
+
);
|
761
|
+
|
762
|
+
bool sample_col(size_t &col_chosen, RNG_engine &rnd_generator);
|
763
|
+
|
764
|
+
void backup(SingleNodeColumnSampler<ldouble_safe, real_t> &other, size_t ncols_tot);
|
765
|
+
|
766
|
+
void restore(const SingleNodeColumnSampler<ldouble_safe, real_t> &other);
|
767
|
+
};
|
768
|
+
|
769
|
+
template <class ImputedData, class ldouble_safe, class real_t>
|
770
|
+
struct WorkerMemory {
|
459
771
|
std::vector<size_t> ix_arr;
|
460
772
|
std::vector<size_t> ix_all;
|
461
773
|
RNG_engine rnd_generator;
|
462
|
-
|
463
|
-
std::uniform_real_distribution<double> rbin;
|
774
|
+
UniformUnitInterval rbin;
|
464
775
|
size_t st;
|
465
776
|
size_t end;
|
466
777
|
size_t st_NA;
|
467
778
|
size_t end_NA;
|
468
779
|
size_t split_ix;
|
469
|
-
|
470
|
-
std::vector<double> weights_arr;
|
780
|
+
hashed_map<size_t, double> weights_map;
|
781
|
+
std::vector<double> weights_arr; /* when not ignoring NAs and when using weights as dty */
|
782
|
+
bool changed_weights; /* when using 'missing_action'='Divide' or density weights */
|
471
783
|
double xmin;
|
472
784
|
double xmax;
|
473
|
-
size_t npresent;
|
785
|
+
size_t npresent; /* 'npresent' and 'ncols_tried' are used interchangeable and for unrelated things */
|
474
786
|
bool unsplittable;
|
475
787
|
std::vector<bool> is_repeated;
|
476
|
-
std::vector<char>
|
477
|
-
size_t ncols_tried;
|
788
|
+
std::vector<signed char> categs;
|
789
|
+
size_t ncols_tried; /* 'npresent' and 'ncols_tried' are used interchangeable and for unrelated things */
|
478
790
|
int ncat_tried;
|
479
|
-
std::vector<
|
480
|
-
|
481
|
-
|
791
|
+
std::vector<double> btree_weights; /* only when using weights for sampling */
|
792
|
+
ColumnSampler<ldouble_safe> col_sampler; /* columns can get eliminated, keep a copy for each thread */
|
793
|
+
SingleNodeColumnSampler<ldouble_safe, real_t> node_col_sampler;
|
794
|
+
SingleNodeColumnSampler<ldouble_safe, real_t> node_col_sampler_backup;
|
482
795
|
|
483
796
|
/* for split criterion */
|
484
797
|
std::vector<double> buffer_dbl;
|
485
798
|
std::vector<size_t> buffer_szt;
|
486
|
-
std::vector<char>
|
799
|
+
std::vector<signed char> buffer_chr;
|
487
800
|
double prob_split_type;
|
801
|
+
ColCriterion col_criterion;
|
488
802
|
GainCriterion criterion;
|
489
803
|
double this_gain;
|
490
804
|
double this_split_point;
|
491
805
|
int this_categ;
|
492
|
-
std::vector<char>
|
806
|
+
std::vector<signed char> this_split_categ;
|
493
807
|
bool determine_split;
|
808
|
+
std::vector<double> imputed_x_buffer;
|
809
|
+
double saved_xmedian;
|
810
|
+
double best_xmedian;
|
811
|
+
int saved_cat_mode;
|
812
|
+
int best_cat_mode;
|
813
|
+
std::vector<size_t> col_indices; /* only for full gain calculation */
|
814
|
+
|
815
|
+
/* for weighted column choices */
|
816
|
+
std::vector<double> node_col_weights;
|
817
|
+
std::vector<double> saved_stat1;
|
818
|
+
std::vector<double> saved_stat2;
|
819
|
+
bool has_saved_stats;
|
820
|
+
double* tree_kurtoses; /* only when mixing 'weight_by_kurt' with 'prob_pick_col*' */
|
494
821
|
|
495
822
|
/* for the extended model */
|
496
823
|
size_t ntry;
|
497
824
|
size_t ntaken;
|
498
825
|
size_t ntaken_best;
|
499
|
-
|
500
|
-
|
826
|
+
size_t ntried;
|
827
|
+
bool try_all;
|
828
|
+
size_t col_chosen; /* also used as placeholder in the single-variable model */
|
501
829
|
ColType col_type;
|
502
830
|
double ext_sd;
|
503
|
-
std::vector<size_t> cols_shuffled;
|
504
831
|
std::vector<double> comb_val;
|
505
832
|
std::vector<size_t> col_take;
|
506
833
|
std::vector<ColType> col_take_type;
|
@@ -510,9 +837,10 @@ typedef struct {
|
|
510
837
|
std::vector<double> ext_fill_val;
|
511
838
|
std::vector<double> ext_fill_new;
|
512
839
|
std::vector<int> chosen_cat;
|
513
|
-
std::vector<std::vector<double>>
|
514
|
-
|
515
|
-
|
840
|
+
std::vector<std::vector<double>> ext_cat_coef;
|
841
|
+
UniformMinusOneToOne coef_unif;
|
842
|
+
StandardNormalDistr coef_norm;
|
843
|
+
std::vector<double> sample_weights; /* when using weights and split criterion */
|
516
844
|
|
517
845
|
/* for similarity/distance calculations */
|
518
846
|
std::vector<double> tmat_sep;
|
@@ -522,9 +850,11 @@ typedef struct {
|
|
522
850
|
|
523
851
|
/* when imputing NAs on-the-fly */
|
524
852
|
std::vector<ImputedData> impute_vec;
|
525
|
-
|
853
|
+
hashed_map<size_t, ImputedData> impute_map;
|
526
854
|
|
527
|
-
|
855
|
+
/* for non-depth scoring metric */
|
856
|
+
DensityCalculator<ldouble_safe, real_t> density_calculator;
|
857
|
+
};
|
528
858
|
|
529
859
|
typedef struct WorkerForSimilarity {
|
530
860
|
std::vector<size_t> ix_arr;
|
@@ -538,55 +868,138 @@ typedef struct WorkerForSimilarity {
|
|
538
868
|
bool assume_full_distr; /* doesn't need to have one copy per worker */
|
539
869
|
} WorkerForSimilarity;
|
540
870
|
|
541
|
-
typedef struct {
|
871
|
+
typedef struct WorkerForPredictCSC {
|
872
|
+
std::vector<size_t> ix_arr;
|
873
|
+
size_t st;
|
874
|
+
size_t end;
|
875
|
+
std::vector<double> comb_val;
|
876
|
+
std::vector<double> weights_arr;
|
877
|
+
std::vector<double> depths;
|
878
|
+
} WorkerForPredictCSC;
|
879
|
+
|
880
|
+
class RecursionState {
|
881
|
+
public:
|
542
882
|
size_t st;
|
543
883
|
size_t st_NA;
|
544
884
|
size_t end_NA;
|
545
885
|
size_t split_ix;
|
546
886
|
size_t end;
|
887
|
+
size_t sampler_pos;
|
888
|
+
size_t n_dropped;
|
889
|
+
bool changed_weights;
|
890
|
+
bool full_state;
|
547
891
|
std::vector<size_t> ix_arr;
|
548
892
|
std::vector<bool> cols_possible;
|
893
|
+
std::vector<double> col_sampler_weights;
|
549
894
|
std::unique_ptr<double[]> weights_arr;
|
550
|
-
|
551
|
-
|
895
|
+
|
896
|
+
RecursionState() = default;
|
897
|
+
template <class WorkerMemory>
|
898
|
+
RecursionState(WorkerMemory &workspace, bool full_state);
|
899
|
+
template <class WorkerMemory>
|
900
|
+
void restore_state(WorkerMemory &workspace);
|
901
|
+
};
|
552
902
|
|
553
903
|
/* Function prototypes */
|
554
904
|
|
555
905
|
/* fit_model.cpp */
|
556
|
-
|
906
|
+
template <class real_t, class sparse_ix, class ldouble_safe>
|
907
|
+
int fit_iforest_internal(
|
908
|
+
IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
909
|
+
real_t numeric_data[], size_t ncols_numeric,
|
910
|
+
int categ_data[], size_t ncols_categ, int ncat[],
|
911
|
+
real_t Xc[], sparse_ix Xc_ind[], sparse_ix Xc_indptr[],
|
912
|
+
size_t ndim, size_t ntry, CoefType coef_type, bool coef_by_prop,
|
913
|
+
real_t sample_weights[], bool with_replacement, bool weight_as_sample,
|
914
|
+
size_t nrows, size_t sample_size, size_t ntrees,
|
915
|
+
size_t max_depth, size_t ncols_per_tree,
|
916
|
+
bool limit_depth, bool penalize_range, bool standardize_data,
|
917
|
+
ScoringMetric scoring_metric, bool fast_bratio,
|
918
|
+
bool standardize_dist, double tmat[],
|
919
|
+
double output_depths[], bool standardize_depth,
|
920
|
+
real_t col_weights[], bool weigh_by_kurt,
|
921
|
+
double prob_pick_by_gain_pl, double prob_pick_by_gain_avg,
|
922
|
+
double prob_pick_by_full_gain, double prob_pick_by_dens,
|
923
|
+
double prob_pick_col_by_range, double prob_pick_col_by_var,
|
924
|
+
double prob_pick_col_by_kurt,
|
925
|
+
double min_gain, MissingAction missing_action,
|
926
|
+
CategSplit cat_split_type, NewCategAction new_cat_action,
|
927
|
+
bool all_perm, Imputer *imputer, size_t min_imp_obs,
|
928
|
+
UseDepthImp depth_imp, WeighImpRows weigh_imp_rows, bool impute_at_fit,
|
929
|
+
uint64_t random_seed, int nthreads);
|
930
|
+
template <class real_t, class sparse_ix>
|
557
931
|
int fit_iforest(IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
558
|
-
|
932
|
+
real_t numeric_data[], size_t ncols_numeric,
|
559
933
|
int categ_data[], size_t ncols_categ, int ncat[],
|
560
|
-
|
934
|
+
real_t Xc[], sparse_ix Xc_ind[], sparse_ix Xc_indptr[],
|
561
935
|
size_t ndim, size_t ntry, CoefType coef_type, bool coef_by_prop,
|
562
|
-
|
563
|
-
size_t nrows, size_t sample_size, size_t ntrees,
|
564
|
-
|
936
|
+
real_t sample_weights[], bool with_replacement, bool weight_as_sample,
|
937
|
+
size_t nrows, size_t sample_size, size_t ntrees,
|
938
|
+
size_t max_depth, size_t ncols_per_tree,
|
939
|
+
bool limit_depth, bool penalize_range, bool standardize_data,
|
940
|
+
ScoringMetric scoring_metric, bool fast_bratio,
|
565
941
|
bool standardize_dist, double tmat[],
|
566
942
|
double output_depths[], bool standardize_depth,
|
567
|
-
|
568
|
-
double
|
569
|
-
double
|
943
|
+
real_t col_weights[], bool weigh_by_kurt,
|
944
|
+
double prob_pick_by_gain_pl, double prob_pick_by_gain_avg,
|
945
|
+
double prob_pick_by_full_gain, double prob_pick_by_dens,
|
946
|
+
double prob_pick_col_by_range, double prob_pick_col_by_var,
|
947
|
+
double prob_pick_col_by_kurt,
|
570
948
|
double min_gain, MissingAction missing_action,
|
571
949
|
CategSplit cat_split_type, NewCategAction new_cat_action,
|
572
950
|
bool all_perm, Imputer *imputer, size_t min_imp_obs,
|
573
951
|
UseDepthImp depth_imp, WeighImpRows weigh_imp_rows, bool impute_at_fit,
|
574
|
-
uint64_t random_seed, bool
|
952
|
+
uint64_t random_seed, bool use_long_double, int nthreads);
|
953
|
+
template <class real_t, class sparse_ix>
|
575
954
|
int add_tree(IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
576
|
-
|
955
|
+
real_t numeric_data[], size_t ncols_numeric,
|
577
956
|
int categ_data[], size_t ncols_categ, int ncat[],
|
578
|
-
|
957
|
+
real_t Xc[], sparse_ix Xc_ind[], sparse_ix Xc_indptr[],
|
579
958
|
size_t ndim, size_t ntry, CoefType coef_type, bool coef_by_prop,
|
580
|
-
|
581
|
-
|
582
|
-
|
583
|
-
|
584
|
-
|
959
|
+
real_t sample_weights[], size_t nrows,
|
960
|
+
size_t max_depth, size_t ncols_per_tree,
|
961
|
+
bool limit_depth, bool penalize_range, bool standardize_data,
|
962
|
+
bool fast_bratio,
|
963
|
+
real_t col_weights[], bool weigh_by_kurt,
|
964
|
+
double prob_pick_by_gain_pl, double prob_pick_by_gain_avg,
|
965
|
+
double prob_pick_by_full_gain, double prob_pick_by_dens,
|
966
|
+
double prob_pick_col_by_range, double prob_pick_col_by_var,
|
967
|
+
double prob_pick_col_by_kurt,
|
585
968
|
double min_gain, MissingAction missing_action,
|
586
969
|
CategSplit cat_split_type, NewCategAction new_cat_action,
|
587
970
|
UseDepthImp depth_imp, WeighImpRows weigh_imp_rows,
|
588
|
-
bool all_perm,
|
971
|
+
bool all_perm, Imputer *imputer, size_t min_imp_obs,
|
972
|
+
TreesIndexer *indexer,
|
973
|
+
real_t ref_numeric_data[], int ref_categ_data[],
|
974
|
+
bool ref_is_col_major, size_t ref_ld_numeric, size_t ref_ld_categ,
|
975
|
+
real_t ref_Xc[], sparse_ix ref_Xc_ind[], sparse_ix ref_Xc_indptr[],
|
976
|
+
uint64_t random_seed, bool use_long_double);
|
977
|
+
template <class real_t, class sparse_ix, class ldouble_safe>
|
978
|
+
int add_tree_internal(
|
979
|
+
IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
980
|
+
real_t numeric_data[], size_t ncols_numeric,
|
981
|
+
int categ_data[], size_t ncols_categ, int ncat[],
|
982
|
+
real_t Xc[], sparse_ix Xc_ind[], sparse_ix Xc_indptr[],
|
983
|
+
size_t ndim, size_t ntry, CoefType coef_type, bool coef_by_prop,
|
984
|
+
real_t sample_weights[], size_t nrows,
|
985
|
+
size_t max_depth, size_t ncols_per_tree,
|
986
|
+
bool limit_depth, bool penalize_range, bool standardize_data,
|
987
|
+
bool fast_bratio,
|
988
|
+
real_t col_weights[], bool weigh_by_kurt,
|
989
|
+
double prob_pick_by_gain_pl, double prob_pick_by_gain_avg,
|
990
|
+
double prob_pick_by_full_gain, double prob_pick_by_dens,
|
991
|
+
double prob_pick_col_by_range, double prob_pick_col_by_var,
|
992
|
+
double prob_pick_col_by_kurt,
|
993
|
+
double min_gain, MissingAction missing_action,
|
994
|
+
CategSplit cat_split_type, NewCategAction new_cat_action,
|
995
|
+
UseDepthImp depth_imp, WeighImpRows weigh_imp_rows,
|
996
|
+
bool all_perm, Imputer *imputer, size_t min_imp_obs,
|
997
|
+
TreesIndexer *indexer,
|
998
|
+
real_t ref_numeric_data[], int ref_categ_data[],
|
999
|
+
bool ref_is_col_major, size_t ref_ld_numeric, size_t ref_ld_categ,
|
1000
|
+
real_t ref_Xc[], sparse_ix ref_Xc_ind[], sparse_ix ref_Xc_indptr[],
|
589
1001
|
uint64_t random_seed);
|
1002
|
+
template <class InputData, class WorkerMemory, class ldouble_safe>
|
590
1003
|
void fit_itree(std::vector<IsoTree> *tree_root,
|
591
1004
|
std::vector<IsoHPlane> *hplane_root,
|
592
1005
|
WorkerMemory &workspace,
|
@@ -596,6 +1009,7 @@ void fit_itree(std::vector<IsoTree> *tree_root,
|
|
596
1009
|
size_t tree_num);
|
597
1010
|
|
598
1011
|
/* isoforest.cpp */
|
1012
|
+
template <class InputData, class WorkerMemory, class ldouble_safe>
|
599
1013
|
void split_itree_recursive(std::vector<IsoTree> &trees,
|
600
1014
|
WorkerMemory &workspace,
|
601
1015
|
InputData &input_data,
|
@@ -604,31 +1018,55 @@ void split_itree_recursive(std::vector<IsoTree> &trees,
|
|
604
1018
|
size_t curr_depth);
|
605
1019
|
|
606
1020
|
/* extended.cpp */
|
1021
|
+
template <class InputData, class WorkerMemory, class ldouble_safe>
|
607
1022
|
void split_hplane_recursive(std::vector<IsoHPlane> &hplanes,
|
608
1023
|
WorkerMemory &workspace,
|
609
1024
|
InputData &input_data,
|
610
1025
|
ModelParams &model_params,
|
611
1026
|
std::vector<ImputeNode> *impute_nodes,
|
612
1027
|
size_t curr_depth);
|
1028
|
+
template <class InputData, class WorkerMemory, class ldouble_safe>
|
613
1029
|
void add_chosen_column(WorkerMemory &workspace, InputData &input_data, ModelParams &model_params,
|
614
|
-
std::vector<bool> &col_is_taken,
|
1030
|
+
std::vector<bool> &col_is_taken, hashed_set<size_t> &col_is_taken_s);
|
615
1031
|
void shrink_to_fit_hplane(IsoHPlane &hplane, bool clear_vectors);
|
1032
|
+
template <class InputData, class WorkerMemory>
|
616
1033
|
void simplify_hplane(IsoHPlane &hplane, WorkerMemory &workspace, InputData &input_data, ModelParams &model_params);
|
617
1034
|
|
618
1035
|
|
619
1036
|
/* predict.cpp */
|
620
|
-
|
621
|
-
|
622
|
-
|
1037
|
+
template <class real_t, class sparse_ix>
|
1038
|
+
#ifndef _FOR_R
|
1039
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno"), gnu::hot]]
|
1040
|
+
#endif
|
1041
|
+
void predict_iforest(real_t *restrict numeric_data, int *restrict categ_data,
|
1042
|
+
bool is_col_major, size_t ld_numeric, size_t ld_categ,
|
1043
|
+
real_t *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1044
|
+
real_t *restrict Xr, sparse_ix *restrict Xr_ind, sparse_ix *restrict Xr_indptr,
|
623
1045
|
size_t nrows, int nthreads, bool standardize,
|
624
1046
|
IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
625
|
-
double output_depths
|
1047
|
+
double *restrict output_depths, sparse_ix *restrict tree_num,
|
1048
|
+
double *restrict per_tree_depths,
|
1049
|
+
TreesIndexer *indexer);
|
1050
|
+
template <class real_t, class sparse_ix>
|
1051
|
+
[[gnu::hot]]
|
1052
|
+
void traverse_itree_fast(std::vector<IsoTree> &tree,
|
1053
|
+
IsoForest &model_outputs,
|
1054
|
+
real_t *restrict row_numeric_data,
|
1055
|
+
double &restrict output_depth,
|
1056
|
+
sparse_ix *restrict tree_num,
|
1057
|
+
double *restrict tree_depth,
|
1058
|
+
size_t row) noexcept;
|
1059
|
+
template <class PredictionData, class sparse_ix>
|
1060
|
+
[[gnu::hot]]
|
626
1061
|
void traverse_itree_no_recurse(std::vector<IsoTree> &tree,
|
627
1062
|
IsoForest &model_outputs,
|
628
1063
|
PredictionData &prediction_data,
|
629
|
-
double
|
1064
|
+
double &restrict output_depth,
|
630
1065
|
sparse_ix *restrict tree_num,
|
631
|
-
|
1066
|
+
double *restrict tree_depth,
|
1067
|
+
size_t row) noexcept;
|
1068
|
+
template <class PredictionData, class sparse_ix, class ImputedData>
|
1069
|
+
[[gnu::hot]]
|
632
1070
|
double traverse_itree(std::vector<IsoTree> &tree,
|
633
1071
|
IsoForest &model_outputs,
|
634
1072
|
PredictionData &prediction_data,
|
@@ -637,63 +1075,181 @@ double traverse_itree(std::vector<IsoTree> &tree,
|
|
637
1075
|
double curr_weight,
|
638
1076
|
size_t row,
|
639
1077
|
sparse_ix *restrict tree_num,
|
640
|
-
|
641
|
-
|
642
|
-
|
643
|
-
|
644
|
-
|
645
|
-
|
646
|
-
|
1078
|
+
double *restrict tree_depth,
|
1079
|
+
size_t curr_lev) noexcept;
|
1080
|
+
template <class PredictionData, class sparse_ix>
|
1081
|
+
[[gnu::hot]]
|
1082
|
+
void traverse_hplane_fast_colmajor(std::vector<IsoHPlane> &hplane,
|
1083
|
+
ExtIsoForest &model_outputs,
|
1084
|
+
PredictionData &prediction_data,
|
1085
|
+
double &restrict output_depth,
|
1086
|
+
sparse_ix *restrict tree_num,
|
1087
|
+
double *restrict tree_depth,
|
1088
|
+
size_t row) noexcept;
|
1089
|
+
template <class real_t, class sparse_ix>
|
1090
|
+
[[gnu::hot]]
|
1091
|
+
void traverse_hplane_fast_rowmajor(std::vector<IsoHPlane> &hplane,
|
1092
|
+
ExtIsoForest &model_outputs,
|
1093
|
+
real_t *restrict row_numeric_data,
|
1094
|
+
double &restrict output_depth,
|
1095
|
+
sparse_ix *restrict tree_num,
|
1096
|
+
double *restrict tree_depth,
|
1097
|
+
size_t row) noexcept;
|
1098
|
+
template <class PredictionData, class sparse_ix, class ImputedData>
|
1099
|
+
[[gnu::hot]]
|
647
1100
|
void traverse_hplane(std::vector<IsoHPlane> &hplane,
|
648
1101
|
ExtIsoForest &model_outputs,
|
649
1102
|
PredictionData &prediction_data,
|
650
|
-
double
|
1103
|
+
double &restrict output_depth,
|
651
1104
|
std::vector<ImputeNode> *impute_nodes,
|
652
1105
|
ImputedData *imputed_data,
|
653
1106
|
sparse_ix *restrict tree_num,
|
654
|
-
|
655
|
-
|
656
|
-
|
657
|
-
void
|
658
|
-
|
1107
|
+
double *restrict tree_depth,
|
1108
|
+
size_t row) noexcept;
|
1109
|
+
template <class real_t, class sparse_ix>
|
1110
|
+
void batched_csc_predict(PredictionData<real_t, sparse_ix> &prediction_data, int nthreads,
|
1111
|
+
IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
1112
|
+
double *restrict output_depths, sparse_ix *restrict tree_num,
|
1113
|
+
double *restrict per_tree_depths);
|
1114
|
+
template <class PredictionData, class sparse_ix>
|
1115
|
+
void traverse_itree_csc(WorkerForPredictCSC &workspace,
|
1116
|
+
std::vector<IsoTree> &trees,
|
1117
|
+
IsoForest &model_outputs,
|
1118
|
+
PredictionData &prediction_data,
|
1119
|
+
sparse_ix *restrict tree_num,
|
1120
|
+
double *restrict per_tree_depths,
|
1121
|
+
size_t curr_tree,
|
1122
|
+
bool has_range_penalty);
|
1123
|
+
template <class PredictionData, class sparse_ix>
|
1124
|
+
void traverse_hplane_csc(WorkerForPredictCSC &workspace,
|
1125
|
+
std::vector<IsoHPlane> &hplanes,
|
1126
|
+
ExtIsoForest &model_outputs,
|
1127
|
+
PredictionData &prediction_data,
|
1128
|
+
sparse_ix *restrict tree_num,
|
1129
|
+
double *restrict per_tree_depths,
|
1130
|
+
size_t curr_tree,
|
1131
|
+
bool has_range_penalty);
|
1132
|
+
template <class PredictionData>
|
1133
|
+
void add_csc_range_penalty(WorkerForPredictCSC &workspace,
|
1134
|
+
PredictionData &prediction_data,
|
1135
|
+
double *restrict weights_arr,
|
1136
|
+
size_t col_num,
|
1137
|
+
double range_low,
|
1138
|
+
double range_high);
|
1139
|
+
template <class PredictionData>
|
1140
|
+
double extract_spC(PredictionData &prediction_data, size_t row, size_t col_num) noexcept;
|
1141
|
+
template <class PredictionData, class sparse_ix>
|
1142
|
+
static inline double extract_spR(PredictionData &prediction_data, sparse_ix *row_st, sparse_ix *row_end, size_t col_num, size_t lb, size_t ub) noexcept;
|
1143
|
+
template <class PredictionData, class sparse_ix>
|
1144
|
+
double extract_spR(PredictionData &prediction_data, sparse_ix *row_st, sparse_ix *row_end, size_t col_num) noexcept;
|
1145
|
+
template <class sparse_ix>
|
1146
|
+
void get_num_nodes(IsoForest &model_outputs, sparse_ix *restrict n_nodes, sparse_ix *restrict n_terminal, int nthreads) noexcept;
|
1147
|
+
template <class sparse_ix>
|
1148
|
+
void get_num_nodes(ExtIsoForest &model_outputs, sparse_ix *restrict n_nodes, sparse_ix *restrict n_terminal, int nthreads) noexcept;
|
659
1149
|
|
660
1150
|
/* dist.cpp */
|
661
|
-
|
662
|
-
|
663
|
-
|
1151
|
+
template <class real_t, class sparse_ix>
|
1152
|
+
void calc_similarity(real_t numeric_data[], int categ_data[],
|
1153
|
+
real_t Xc[], sparse_ix Xc_ind[], sparse_ix Xc_indptr[],
|
1154
|
+
size_t nrows, bool use_long_double, int nthreads,
|
1155
|
+
bool assume_full_distr, bool standardize_dist, bool as_kernel,
|
664
1156
|
IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
665
|
-
double tmat[], double rmat[], size_t n_from
|
1157
|
+
double tmat[], double rmat[], size_t n_from, bool use_indexed_references,
|
1158
|
+
TreesIndexer *indexer, bool is_col_major, size_t ld_numeric, size_t ld_categ);
|
1159
|
+
template <class real_t, class sparse_ix, class ldouble_safe>
|
1160
|
+
void calc_similarity_internal(
|
1161
|
+
real_t numeric_data[], int categ_data[],
|
1162
|
+
real_t Xc[], sparse_ix Xc_ind[], sparse_ix Xc_indptr[],
|
1163
|
+
size_t nrows, int nthreads,
|
1164
|
+
bool assume_full_distr, bool standardize_dist, bool as_kernel,
|
1165
|
+
IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
1166
|
+
double tmat[], double rmat[], size_t n_from, bool use_indexed_references,
|
1167
|
+
TreesIndexer *indexer, bool is_col_major, size_t ld_numeric, size_t ld_categ);
|
1168
|
+
template <class PredictionData, class ldouble_safe>
|
666
1169
|
void traverse_tree_sim(WorkerForSimilarity &workspace,
|
667
1170
|
PredictionData &prediction_data,
|
668
1171
|
IsoForest &model_outputs,
|
669
1172
|
std::vector<IsoTree> &trees,
|
670
|
-
size_t curr_tree
|
1173
|
+
size_t curr_tree,
|
1174
|
+
const bool as_kernel);
|
1175
|
+
template <class PredictionData, class ldouble_safe>
|
671
1176
|
void traverse_hplane_sim(WorkerForSimilarity &workspace,
|
672
1177
|
PredictionData &prediction_data,
|
673
1178
|
ExtIsoForest &model_outputs,
|
674
1179
|
std::vector<IsoHPlane> &hplanes,
|
675
|
-
size_t curr_tree
|
1180
|
+
size_t curr_tree,
|
1181
|
+
const bool as_kernel);
|
1182
|
+
template <class PredictionData, class InputData, class WorkerMemory>
|
1183
|
+
#ifndef _FOR_R
|
1184
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
1185
|
+
#endif
|
676
1186
|
void gather_sim_result(std::vector<WorkerForSimilarity> *worker_memory,
|
677
1187
|
std::vector<WorkerMemory> *worker_memory_m,
|
678
1188
|
PredictionData *prediction_data, InputData *input_data,
|
679
1189
|
IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
680
1190
|
double *restrict tmat, double *restrict rmat, size_t n_from,
|
681
1191
|
size_t ntrees, bool assume_full_distr,
|
682
|
-
bool standardize_dist, int nthreads);
|
1192
|
+
bool standardize_dist, bool as_kernel, int nthreads);
|
1193
|
+
template <class PredictionData>
|
683
1194
|
void initialize_worker_for_sim(WorkerForSimilarity &workspace,
|
684
1195
|
PredictionData &prediction_data,
|
685
1196
|
IsoForest *model_outputs,
|
686
1197
|
ExtIsoForest *model_outputs_ext,
|
687
1198
|
size_t n_from,
|
688
1199
|
bool assume_full_distr);
|
1200
|
+
template <class real_t, class sparse_ix>
|
1201
|
+
#ifndef _FOR_R
|
1202
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
1203
|
+
#endif
|
1204
|
+
void calc_similarity_from_indexer
|
1205
|
+
(
|
1206
|
+
real_t *restrict numeric_data, int *restrict categ_data,
|
1207
|
+
real_t *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1208
|
+
size_t nrows, int nthreads, bool assume_full_distr, bool standardize_dist,
|
1209
|
+
IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
1210
|
+
double *restrict tmat, double *restrict rmat, size_t n_from,
|
1211
|
+
TreesIndexer *indexer, bool is_col_major, size_t ld_numeric, size_t ld_categ
|
1212
|
+
);
|
1213
|
+
template <class real_t, class sparse_ix>
|
1214
|
+
#ifndef _FOR_R
|
1215
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
1216
|
+
#endif
|
1217
|
+
void calc_similarity_from_indexer_with_references
|
1218
|
+
(
|
1219
|
+
real_t *restrict numeric_data, int *restrict categ_data,
|
1220
|
+
real_t *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1221
|
+
size_t nrows, int nthreads, bool standardize_dist,
|
1222
|
+
IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
1223
|
+
double *restrict rmat,
|
1224
|
+
TreesIndexer *indexer, bool is_col_major, size_t ld_numeric, size_t ld_categ
|
1225
|
+
);
|
1226
|
+
template <class real_t, class sparse_ix>
|
1227
|
+
void kernel_to_references(TreesIndexer &indexer,
|
1228
|
+
IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
1229
|
+
real_t *restrict numeric_data, int *restrict categ_data,
|
1230
|
+
real_t *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1231
|
+
bool is_col_major, size_t ld_numeric, size_t ld_categ,
|
1232
|
+
size_t nrows, int nthreads,
|
1233
|
+
double *restrict rmat,
|
1234
|
+
bool standardize);
|
689
1235
|
|
690
1236
|
/* impute.cpp */
|
691
|
-
|
692
|
-
|
1237
|
+
template <class real_t, class sparse_ix>
|
1238
|
+
void impute_missing_values(real_t numeric_data[], int categ_data[], bool is_col_major,
|
1239
|
+
real_t Xr[], sparse_ix Xr_ind[], sparse_ix Xr_indptr[],
|
1240
|
+
size_t nrows, bool use_long_double, int nthreads,
|
1241
|
+
IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
1242
|
+
Imputer &imputer);
|
1243
|
+
template <class real_t, class sparse_ix, class ldouble_safe>
|
1244
|
+
void impute_missing_values_internal(
|
1245
|
+
real_t numeric_data[], int categ_data[], bool is_col_major,
|
1246
|
+
real_t Xr[], sparse_ix Xr_ind[], sparse_ix Xr_indptr[],
|
693
1247
|
size_t nrows, int nthreads,
|
694
1248
|
IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
695
1249
|
Imputer &imputer);
|
1250
|
+
template <class InputData, class ldouble_safe>
|
696
1251
|
void initialize_imputer(Imputer &imputer, InputData &input_data, size_t ntrees, int nthreads);
|
1252
|
+
template <class InputData, class WorkerMemory, class ldouble_safe>
|
697
1253
|
void build_impute_node(ImputeNode &imputer, WorkerMemory &workspace,
|
698
1254
|
InputData &input_data, ModelParams &model_params,
|
699
1255
|
std::vector<ImputeNode> &imputer_tree,
|
@@ -702,232 +1258,1085 @@ void shrink_impute_node(ImputeNode &imputer);
|
|
702
1258
|
void drop_nonterminal_imp_node(std::vector<ImputeNode> &imputer_tree,
|
703
1259
|
std::vector<IsoTree> *trees,
|
704
1260
|
std::vector<IsoHPlane> *hplanes);
|
705
|
-
|
1261
|
+
template <class ImputedData>
|
1262
|
+
void combine_imp_single(ImputedData &restrict imp_addfrom, ImputedData &restrict imp_addto);
|
1263
|
+
template <class ImputedData, class WorkerMemory>
|
706
1264
|
void combine_tree_imputations(WorkerMemory &workspace,
|
707
1265
|
std::vector<ImputedData> &impute_vec,
|
708
|
-
|
1266
|
+
hashed_map<size_t, ImputedData> &impute_map,
|
709
1267
|
std::vector<char> &has_missing,
|
710
1268
|
int nthreads);
|
1269
|
+
template <class ImputedData>
|
711
1270
|
void add_from_impute_node(ImputeNode &imputer, ImputedData &imputed_data, double w);
|
1271
|
+
template <class InputData, class WorkerMemory>
|
712
1272
|
void add_from_impute_node(ImputeNode &imputer, WorkerMemory &workspace, InputData &input_data);
|
713
|
-
template <class imp_arr>
|
1273
|
+
template <class imp_arr, class InputData>
|
714
1274
|
void apply_imputation_results(imp_arr &impute_vec,
|
715
1275
|
Imputer &imputer,
|
716
1276
|
InputData &input_data,
|
717
1277
|
int nthreads);
|
1278
|
+
template <class ImputedData, class InputData>
|
718
1279
|
void apply_imputation_results(std::vector<ImputedData> &impute_vec,
|
719
|
-
|
1280
|
+
hashed_map<size_t, ImputedData> &impute_map,
|
720
1281
|
Imputer &imputer,
|
721
1282
|
InputData &input_data,
|
722
1283
|
int nthreads);
|
1284
|
+
template <class PredictionData, class ImputedData>
|
723
1285
|
void apply_imputation_results(PredictionData &prediction_data,
|
724
1286
|
ImputedData &imp,
|
725
1287
|
Imputer &imputer,
|
726
1288
|
size_t row);
|
1289
|
+
template <class ImputedData, class InputData>
|
727
1290
|
void initialize_impute_calc(ImputedData &imp, InputData &input_data, size_t row);
|
1291
|
+
template <class ImputedData, class PredictionData>
|
728
1292
|
void initialize_impute_calc(ImputedData &imp, PredictionData &prediction_data, Imputer &imputer, size_t row);
|
1293
|
+
template <class ImputedData, class InputData>
|
729
1294
|
void allocate_imp_vec(std::vector<ImputedData> &impute_vec, InputData &input_data, int nthreads);
|
730
|
-
|
1295
|
+
template <class ImputedData, class InputData>
|
1296
|
+
void allocate_imp_map(hashed_map<size_t, ImputedData> &impute_map, InputData &input_data);
|
1297
|
+
template <class ImputedData, class InputData>
|
731
1298
|
void allocate_imp(InputData &input_data,
|
732
1299
|
std::vector<ImputedData> &impute_vec,
|
733
|
-
|
1300
|
+
hashed_map<size_t, ImputedData> &impute_map,
|
734
1301
|
int nthreads);
|
1302
|
+
template <class ImputedData, class InputData>
|
735
1303
|
void check_for_missing(InputData &input_data,
|
736
1304
|
std::vector<ImputedData> &impute_vec,
|
737
|
-
|
1305
|
+
hashed_map<size_t, ImputedData> &impute_map,
|
738
1306
|
int nthreads);
|
1307
|
+
template <class PredictionData>
|
739
1308
|
size_t check_for_missing(PredictionData &prediction_data,
|
740
1309
|
Imputer &imputer,
|
741
1310
|
size_t ix_arr[],
|
742
1311
|
int nthreads);
|
743
1312
|
|
744
1313
|
/* helpers_iforest.cpp */
|
745
|
-
|
746
|
-
|
747
|
-
|
748
|
-
|
749
|
-
|
750
|
-
|
1314
|
+
static inline size_t get_ntrees(const IsoForest &model)
|
1315
|
+
{
|
1316
|
+
return model.trees.size();
|
1317
|
+
}
|
1318
|
+
|
1319
|
+
static inline size_t get_ntrees(const ExtIsoForest &model)
|
1320
|
+
{
|
1321
|
+
return model.hplanes.size();
|
1322
|
+
}
|
1323
|
+
|
1324
|
+
static inline size_t get_ntrees(const Imputer &model)
|
1325
|
+
{
|
1326
|
+
return model.imputer_tree.size();
|
1327
|
+
}
|
1328
|
+
|
1329
|
+
static inline size_t get_ntrees(const TreesIndexer &model)
|
1330
|
+
{
|
1331
|
+
return model.indices.size();
|
1332
|
+
}
|
1333
|
+
template <class InputData, class WorkerMemory>
|
751
1334
|
void get_split_range(WorkerMemory &workspace, InputData &input_data, ModelParams &model_params, IsoTree &tree);
|
1335
|
+
template <class InputData, class WorkerMemory>
|
752
1336
|
void get_split_range(WorkerMemory &workspace, InputData &input_data, ModelParams &model_params);
|
1337
|
+
template <class InputData, class WorkerMemory>
|
1338
|
+
void get_split_range_v2(WorkerMemory &workspace, InputData &input_data, ModelParams &model_params);
|
1339
|
+
template <class InputData, class WorkerMemory>
|
753
1340
|
int choose_cat_from_present(WorkerMemory &workspace, InputData &input_data, size_t col_num);
|
754
|
-
|
755
|
-
|
756
|
-
|
757
|
-
void set_col_as_taken(std::vector<bool> &col_is_taken,
|
1341
|
+
bool is_col_taken(std::vector<bool> &col_is_taken, hashed_set<size_t> &col_is_taken_s,
|
1342
|
+
size_t col_num);
|
1343
|
+
template <class InputData>
|
1344
|
+
void set_col_as_taken(std::vector<bool> &col_is_taken, hashed_set<size_t> &col_is_taken_s,
|
758
1345
|
InputData &input_data, size_t col_num, ColType col_type);
|
1346
|
+
template <class InputData>
|
1347
|
+
void set_col_as_taken(std::vector<bool> &col_is_taken, hashed_set<size_t> &col_is_taken_s,
|
1348
|
+
InputData &input_data, size_t col_num);
|
1349
|
+
template <class InputData, class WorkerMemory>
|
759
1350
|
void add_separation_step(WorkerMemory &workspace, InputData &input_data, double remainder);
|
760
|
-
|
1351
|
+
template <class InputData, class WorkerMemory, class ldouble_safe>
|
1352
|
+
void add_remainder_separation_steps(WorkerMemory &workspace, InputData &input_data, ldouble_safe sum_weight);
|
1353
|
+
template <class PredictionData, class sparse_ix>
|
761
1354
|
void remap_terminal_trees(IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
762
1355
|
PredictionData &prediction_data, sparse_ix *restrict tree_num, int nthreads);
|
763
|
-
|
764
|
-
|
1356
|
+
template <class InputData, class ldouble_safe>
|
1357
|
+
std::vector<double> calc_kurtosis_all_data(InputData &input_data, ModelParams &model_params, RNG_engine &rnd_generator);
|
1358
|
+
template <class InputData, class WorkerMemory>
|
1359
|
+
void calc_ranges_all_cols(InputData &input_data, WorkerMemory &workspace, ModelParams &model_params,
|
1360
|
+
double *restrict ranges, double *restrict saved_xmin, double *restrict saved_xmax);
|
1361
|
+
template <class InputData, class WorkerMemory, class ldouble_safe>
|
1362
|
+
void calc_var_all_cols(InputData &input_data, WorkerMemory &workspace, ModelParams &model_params,
|
1363
|
+
double *restrict variances, double *restrict saved_xmin, double *restrict saved_xmax,
|
1364
|
+
double *restrict saved_means, double *restrict saved_sds);
|
1365
|
+
template <class InputData, class WorkerMemory, class ldouble_safe>
|
1366
|
+
void calc_kurt_all_cols(InputData &input_data, WorkerMemory &workspace, ModelParams &model_params,
|
1367
|
+
double *restrict kurtosis, double *restrict saved_xmin, double *restrict saved_xmax);
|
1368
|
+
bool is_boxed_metric(const ScoringMetric scoring_metric);
|
765
1369
|
|
766
1370
|
|
767
1371
|
/* utils.cpp */
|
1372
|
+
#define ix_comb_(i, j, n, ncomb) ( ((ncomb) + ((j) - (i))) - (size_t)1 - div2(((n) - (i)) * ((n) - (i) - (size_t)1)) )
|
1373
|
+
#define ix_comb(i, j, n, ncomb) ( ((i) < (j))? ix_comb_(i, j, n, ncomb) : ix_comb_(j, i, n, ncomb) )
|
1374
|
+
#define calc_ncomb(n) (((n) % 2) == 0)? (div2(n) * ((n)-(size_t)1)) : ((n) * div2((n)-(size_t)1))
|
768
1375
|
size_t log2ceil(size_t x);
|
1376
|
+
#ifndef _FOR_R
|
1377
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
1378
|
+
#endif
|
1379
|
+
double digamma(double x);
|
1380
|
+
template <class ldouble_safe>
|
1381
|
+
#ifndef _FOR_R
|
1382
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
1383
|
+
#endif
|
769
1384
|
double harmonic(size_t n);
|
770
1385
|
double harmonic_recursive(double a, double b);
|
1386
|
+
template <class ldouble_safe>
|
771
1387
|
double expected_avg_depth(size_t sample_size);
|
772
|
-
|
1388
|
+
template <class ldouble_safe>
|
1389
|
+
#ifndef _FOR_R
|
1390
|
+
[[gnu::optimize("no-trapping-math"), gnu::optimize("no-math-errno")]]
|
1391
|
+
#endif
|
1392
|
+
double expected_avg_depth(ldouble_safe approx_sample_size);
|
773
1393
|
double expected_separation_depth(size_t n);
|
774
1394
|
double expected_separation_depth_hotstart(double curr, size_t n_curr, size_t n_final);
|
775
|
-
|
1395
|
+
template <class ldouble_safe>
|
1396
|
+
double expected_separation_depth(ldouble_safe n);
|
776
1397
|
void increase_comb_counter(size_t ix_arr[], size_t st, size_t end, size_t n, double counter[], double exp_remainder);
|
777
1398
|
void increase_comb_counter(size_t ix_arr[], size_t st, size_t end, size_t n,
|
778
1399
|
double *restrict counter, double *restrict weights, double exp_remainder);
|
779
1400
|
void increase_comb_counter(size_t ix_arr[], size_t st, size_t end, size_t n,
|
780
|
-
double counter[],
|
1401
|
+
double counter[], hashed_map<size_t, double> &weights, double exp_remainder);
|
781
1402
|
void increase_comb_counter_in_groups(size_t ix_arr[], size_t st, size_t end, size_t split_ix, size_t n,
|
782
1403
|
double counter[], double exp_remainder);
|
783
1404
|
void increase_comb_counter_in_groups(size_t ix_arr[], size_t st, size_t end, size_t split_ix, size_t n,
|
784
1405
|
double *restrict counter, double *restrict weights, double exp_remainder);
|
785
|
-
void tmat_to_dense(double *restrict tmat, double *restrict dmat, size_t n,
|
786
|
-
|
787
|
-
|
788
|
-
|
789
|
-
|
790
|
-
void sample_random_rows(std::vector<size_t> &ix_arr, size_t nrows, bool with_replacement,
|
791
|
-
RNG_engine &rnd_generator, std::vector<size_t> &ix_all,
|
792
|
-
|
1406
|
+
void tmat_to_dense(double *restrict tmat, double *restrict dmat, size_t n, double fill_diag);
|
1407
|
+
template <class real_t=double>
|
1408
|
+
void build_btree_sampler(std::vector<double> &btree_weights, real_t *restrict sample_weights,
|
1409
|
+
size_t nrows, size_t &restrict log2_n, size_t &restrict btree_offset);
|
1410
|
+
template <class real_t=double, class ldouble_safe>
|
1411
|
+
void sample_random_rows(std::vector<size_t> &restrict ix_arr, size_t nrows, bool with_replacement,
|
1412
|
+
RNG_engine &rnd_generator, std::vector<size_t> &restrict ix_all,
|
1413
|
+
real_t *restrict sample_weights, std::vector<double> &restrict btree_weights,
|
793
1414
|
size_t log2_n, size_t btree_offset, std::vector<bool> &is_repeated);
|
794
|
-
|
795
|
-
|
796
|
-
|
797
|
-
|
798
|
-
|
799
|
-
|
800
|
-
MissingAction missing_action, size_t &st_NA, size_t &end_NA, size_t &split_ix);
|
801
|
-
|
802
|
-
|
803
|
-
|
1415
|
+
template <class real_t=double>
|
1416
|
+
void weighted_shuffle(size_t *restrict outp, size_t n, real_t *restrict weights, double *restrict buffer_arr, RNG_engine &rnd_generator);
|
1417
|
+
double sample_random_uniform(double xmin, double xmax, RNG_engine &rng) noexcept;
|
1418
|
+
size_t divide_subset_split(size_t ix_arr[], double x[], size_t st, size_t end, double split_point) noexcept;
|
1419
|
+
template <class real_t=double>
|
1420
|
+
void divide_subset_split(size_t *restrict ix_arr, real_t x[], size_t st, size_t end, double split_point,
|
1421
|
+
MissingAction missing_action, size_t &restrict st_NA, size_t &restrict end_NA, size_t &restrict split_ix) noexcept;
|
1422
|
+
template <class real_t, class sparse_ix>
|
1423
|
+
void divide_subset_split(size_t *restrict ix_arr, size_t st, size_t end, size_t col_num,
|
1424
|
+
real_t Xc[], sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr, double split_point,
|
1425
|
+
MissingAction missing_action, size_t &restrict st_NA, size_t &restrict end_NA, size_t &restrict split_ix) noexcept;
|
1426
|
+
void divide_subset_split(size_t *restrict ix_arr, int x[], size_t st, size_t end, signed char split_categ[],
|
1427
|
+
MissingAction missing_action, size_t &restrict st_NA, size_t &restrict end_NA, size_t &restrict split_ix) noexcept;
|
1428
|
+
void divide_subset_split(size_t *restrict ix_arr, int x[], size_t st, size_t end, signed char split_categ[],
|
804
1429
|
int ncat, MissingAction missing_action, NewCategAction new_cat_action,
|
805
|
-
bool move_new_to_left, size_t &st_NA, size_t &end_NA, size_t &split_ix);
|
806
|
-
void divide_subset_split(size_t ix_arr
|
807
|
-
MissingAction missing_action, size_t &st_NA, size_t &end_NA, size_t &split_ix);
|
808
|
-
void divide_subset_split(size_t ix_arr
|
1430
|
+
bool move_new_to_left, size_t &restrict st_NA, size_t &restrict end_NA, size_t &restrict split_ix) noexcept;
|
1431
|
+
void divide_subset_split(size_t *restrict ix_arr, int x[], size_t st, size_t end, int split_categ,
|
1432
|
+
MissingAction missing_action, size_t &restrict st_NA, size_t &restrict end_NA, size_t &restrict split_ix) noexcept;
|
1433
|
+
void divide_subset_split(size_t *restrict ix_arr, int x[], size_t st, size_t end,
|
809
1434
|
MissingAction missing_action, NewCategAction new_cat_action,
|
810
|
-
bool move_new_to_left, size_t &st_NA, size_t &end_NA, size_t &split_ix);
|
811
|
-
|
812
|
-
|
813
|
-
|
814
|
-
|
815
|
-
|
816
|
-
|
817
|
-
|
818
|
-
|
819
|
-
|
1435
|
+
bool move_new_to_left, size_t &restrict st_NA, size_t &restrict end_NA, size_t &restrict split_ix) noexcept;
|
1436
|
+
template <class real_t=double>
|
1437
|
+
void get_range(size_t ix_arr[], real_t *restrict x, size_t st, size_t end,
|
1438
|
+
MissingAction missing_action, double &restrict xmin, double &restrict xmax, bool &unsplittable) noexcept;
|
1439
|
+
template <class real_t>
|
1440
|
+
void get_range(real_t *restrict x, size_t n,
|
1441
|
+
MissingAction missing_action, double &restrict xmin, double &restrict xmax, bool &unsplittable) noexcept;
|
1442
|
+
template <class real_t, class sparse_ix>
|
1443
|
+
void get_range(size_t *restrict ix_arr, size_t st, size_t end, size_t col_num,
|
1444
|
+
real_t *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1445
|
+
MissingAction missing_action, double &restrict xmin_, double &restrict xmax_, bool &unsplittable) noexcept;
|
1446
|
+
template <class real_t, class sparse_ix>
|
1447
|
+
void get_range(size_t col_num, size_t nrows,
|
1448
|
+
real_t *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1449
|
+
MissingAction missing_action, double &restrict xmin, double &restrict xmax, bool &unsplittable) noexcept;
|
1450
|
+
void get_categs(size_t *restrict ix_arr, int x[], size_t st, size_t end, int ncat,
|
1451
|
+
MissingAction missing_action, signed char categs[], size_t &restrict npresent, bool &unsplittable) noexcept;
|
1452
|
+
template <class real_t>
|
1453
|
+
bool check_more_than_two_unique_values(size_t ix_arr[], size_t st, size_t end, real_t x[], MissingAction missing_action);
|
1454
|
+
bool check_more_than_two_unique_values(size_t ix_arr[], size_t st, size_t end, int x[], MissingAction missing_action);
|
1455
|
+
template <class real_t, class sparse_ix>
|
1456
|
+
bool check_more_than_two_unique_values(size_t *restrict ix_arr, size_t st, size_t end, size_t col,
|
1457
|
+
sparse_ix *restrict Xc_indptr, sparse_ix *restrict Xc_ind, real_t *restrict Xc,
|
1458
|
+
MissingAction missing_action);
|
1459
|
+
template <class real_t, class sparse_ix>
|
1460
|
+
bool check_more_than_two_unique_values(size_t nrows, size_t col,
|
1461
|
+
sparse_ix *restrict Xc_indptr, sparse_ix *restrict Xc_ind, real_t *restrict Xc,
|
1462
|
+
MissingAction missing_action);
|
1463
|
+
void count_categs(size_t *restrict ix_arr, size_t st, size_t end, int x[], int ncat, size_t *restrict counts);
|
1464
|
+
int count_ncateg_in_col(const int x[], const size_t n, const int ncat, unsigned char buffer[]);
|
1465
|
+
template <class ldouble_safe>
|
1466
|
+
ldouble_safe calculate_sum_weights(std::vector<size_t> &ix_arr, size_t st, size_t end, size_t curr_depth,
|
1467
|
+
std::vector<double> &weights_arr, hashed_map<size_t, double> &weights_map);
|
1468
|
+
extern bool interrupt_switch;
|
1469
|
+
extern bool signal_is_locked;
|
820
1470
|
void set_interrup_global_variable(int s);
|
1471
|
+
#ifdef _FOR_PYTHON
|
1472
|
+
bool cy_check_interrupt_switch();
|
1473
|
+
void cy_tick_off_interrupt_switch();
|
1474
|
+
#endif
|
1475
|
+
class SignalSwitcher
|
1476
|
+
{
|
1477
|
+
public:
|
1478
|
+
sig_t_ old_sig;
|
1479
|
+
bool is_active;
|
1480
|
+
SignalSwitcher();
|
1481
|
+
~SignalSwitcher();
|
1482
|
+
void restore_handle();
|
1483
|
+
};
|
1484
|
+
void check_interrupt_switch(SignalSwitcher &ss);
|
1485
|
+
bool has_long_double();
|
821
1486
|
int return_EXIT_SUCCESS();
|
822
1487
|
int return_EXIT_FAILURE();
|
823
1488
|
|
824
1489
|
|
825
1490
|
|
826
|
-
|
827
|
-
size_t move_NAs_to_front(size_t ix_arr[], size_t st, size_t end,
|
1491
|
+
template <class real_t=double>
|
1492
|
+
size_t move_NAs_to_front(size_t ix_arr[], size_t st, size_t end, real_t x[]);
|
1493
|
+
template <class real_t, class sparse_ix>
|
1494
|
+
size_t move_NAs_to_front(size_t *restrict ix_arr, size_t st, size_t end, size_t col_num, real_t Xc[], sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr);
|
828
1495
|
size_t move_NAs_to_front(size_t ix_arr[], size_t st, size_t end, int x[]);
|
829
|
-
size_t center_NAs(size_t
|
830
|
-
|
831
|
-
|
1496
|
+
size_t center_NAs(size_t ix_arr[], size_t st_left, size_t st, size_t curr_pos);
|
1497
|
+
template <class real_t>
|
1498
|
+
void fill_NAs_with_median(size_t *restrict ix_arr, size_t st_orig, size_t st, size_t end, real_t *restrict x,
|
1499
|
+
double *restrict buffer_imputed_x, double *restrict xmedian);
|
1500
|
+
template <class real_t, class sparse_ix>
|
1501
|
+
void todense(size_t *restrict ix_arr, size_t st, size_t end,
|
1502
|
+
size_t col_num, real_t *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
832
1503
|
double *restrict buffer_arr);
|
1504
|
+
template <class real_t>
|
1505
|
+
void colmajor_to_rowmajor(real_t *restrict X, size_t nrows, size_t ncols, std::vector<double> &X_row_major);
|
1506
|
+
template <class real_t, class sparse_ix>
|
1507
|
+
void colmajor_to_rowmajor(real_t *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1508
|
+
size_t nrows, size_t ncols,
|
1509
|
+
std::vector<double> &Xr, std::vector<size_t> &Xr_ind, std::vector<size_t> &Xr_indptr);
|
1510
|
+
template <class sparse_ix=size_t>
|
1511
|
+
bool check_indices_are_sorted(sparse_ix indices[], size_t n);
|
1512
|
+
template <class real_t, class sparse_ix>
|
1513
|
+
void sort_csc_indices(real_t *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr, size_t ncols_numeric);
|
833
1514
|
|
834
1515
|
/* mult.cpp */
|
835
|
-
|
836
|
-
|
837
|
-
|
838
|
-
|
839
|
-
|
1516
|
+
template <class real_t, class real_t_>
|
1517
|
+
void calc_mean_and_sd_t(size_t ix_arr[], size_t st, size_t end, real_t_ *restrict x,
|
1518
|
+
MissingAction missing_action, double &restrict x_sd, double &restrict x_mean);
|
1519
|
+
template <class real_t_, class ldouble_safe>
|
1520
|
+
void calc_mean_and_sd(size_t ix_arr[], size_t st, size_t end, real_t_ *restrict x,
|
1521
|
+
MissingAction missing_action, double &restrict x_sd, double &restrict x_mean);
|
1522
|
+
template <class real_t_>
|
1523
|
+
double calc_mean_only(size_t ix_arr[], size_t st, size_t end, real_t_ *restrict x);
|
1524
|
+
template <class real_t_, class mapping, class ldouble_safe>
|
1525
|
+
void calc_mean_and_sd_weighted(size_t ix_arr[], size_t st, size_t end, real_t_ *restrict x, mapping &restrict w,
|
1526
|
+
MissingAction missing_action, double &restrict x_sd, double &restrict x_mean);
|
1527
|
+
template <class real_t_, class mapping>
|
1528
|
+
double calc_mean_only_weighted(size_t ix_arr[], size_t st, size_t end, real_t_ *restrict x, mapping &restrict w);
|
1529
|
+
template <class real_t_, class sparse_ix, class ldouble_safe>
|
1530
|
+
void calc_mean_and_sd(size_t *restrict ix_arr, size_t st, size_t end, size_t col_num,
|
1531
|
+
real_t_ *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1532
|
+
double &restrict x_sd, double &restrict x_mean);
|
1533
|
+
template <class real_t_, class sparse_ix, class ldouble_safe>
|
1534
|
+
double calc_mean_only(size_t *restrict ix_arr, size_t st, size_t end, size_t col_num,
|
1535
|
+
real_t_ *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr);
|
1536
|
+
template <class real_t_, class sparse_ix, class mapping, class ldouble_safe>
|
1537
|
+
void calc_mean_and_sd_weighted(size_t *restrict ix_arr, size_t st, size_t end, size_t col_num,
|
1538
|
+
real_t_ *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1539
|
+
double &restrict x_sd, double &restrict x_mean, mapping &restrict w);
|
1540
|
+
template <class real_t_, class sparse_ix, class mapping, class ldouble_safe>
|
1541
|
+
double calc_mean_only_weighted(size_t *restrict ix_arr, size_t st, size_t end, size_t col_num,
|
1542
|
+
real_t_ *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1543
|
+
mapping &restrict w);
|
1544
|
+
template <class real_t_>
|
840
1545
|
void add_linear_comb(size_t ix_arr[], size_t st, size_t end, double *restrict res,
|
841
|
-
|
1546
|
+
real_t_ *restrict x, double &restrict coef, double x_sd, double x_mean, double &restrict fill_val,
|
842
1547
|
MissingAction missing_action, double *restrict buffer_arr,
|
843
1548
|
size_t *restrict buffer_NAs, bool first_run);
|
1549
|
+
template <class real_t_, class mapping, class ldouble_safe>
|
1550
|
+
void add_linear_comb_weighted(size_t ix_arr[], size_t st, size_t end, double *restrict res,
|
1551
|
+
real_t_ *restrict x, double &restrict coef, double x_sd, double x_mean, double &restrict fill_val,
|
1552
|
+
MissingAction missing_action, double *restrict buffer_arr,
|
1553
|
+
size_t *restrict buffer_NAs, bool first_run, mapping &restrict w);
|
1554
|
+
template <class real_t_, class sparse_ix>
|
844
1555
|
void add_linear_comb(size_t *restrict ix_arr, size_t st, size_t end, size_t col_num, double *restrict res,
|
845
|
-
|
846
|
-
double &coef, double x_sd, double x_mean, double &fill_val, MissingAction missing_action,
|
1556
|
+
real_t_ *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1557
|
+
double &restrict coef, double x_sd, double x_mean, double &restrict fill_val, MissingAction missing_action,
|
847
1558
|
double *restrict buffer_arr, size_t *restrict buffer_NAs, bool first_run);
|
1559
|
+
template <class real_t_, class sparse_ix, class mapping, class ldouble_safe>
|
1560
|
+
void add_linear_comb_weighted(size_t *restrict ix_arr, size_t st, size_t end, size_t col_num, double *restrict res,
|
1561
|
+
real_t_ *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1562
|
+
double &restrict coef, double x_sd, double x_mean, double &restrict fill_val, MissingAction missing_action,
|
1563
|
+
double *restrict buffer_arr, size_t *restrict buffer_NAs, bool first_run, mapping &restrict w);
|
1564
|
+
template <class mapping>
|
1565
|
+
void add_linear_comb_weighted(size_t *restrict ix_arr, size_t st, size_t end, double *restrict res,
|
1566
|
+
int x[], int ncat, double *restrict cat_coef, double single_cat_coef, int chosen_cat,
|
1567
|
+
double &restrict fill_val, double &restrict fill_new, size_t *restrict buffer_pos,
|
1568
|
+
NewCategAction new_cat_action, MissingAction missing_action, CategSplit cat_split_type,
|
1569
|
+
bool first_run, mapping &restrict w);
|
1570
|
+
template <class ldouble_safe>
|
848
1571
|
void add_linear_comb(size_t *restrict ix_arr, size_t st, size_t end, double *restrict res,
|
849
1572
|
int x[], int ncat, double *restrict cat_coef, double single_cat_coef, int chosen_cat,
|
850
|
-
double &fill_val, double &fill_new, size_t *restrict buffer_cnt, size_t *restrict buffer_pos,
|
1573
|
+
double &restrict fill_val, double &restrict fill_new, size_t *restrict buffer_cnt, size_t *restrict buffer_pos,
|
851
1574
|
NewCategAction new_cat_action, MissingAction missing_action, CategSplit cat_split_type, bool first_run);
|
1575
|
+
template <class mapping, class ldouble_safe>
|
1576
|
+
void add_linear_comb_weighted(size_t *restrict ix_arr, size_t st, size_t end, double *restrict res,
|
1577
|
+
int x[], int ncat, double *restrict cat_coef, double single_cat_coef, int chosen_cat,
|
1578
|
+
double &restrict fill_val, double &restrict fill_new, size_t *restrict buffer_pos,
|
1579
|
+
NewCategAction new_cat_action, MissingAction missing_action, CategSplit cat_split_type,
|
1580
|
+
bool first_run, mapping &restrict w);
|
852
1581
|
|
853
1582
|
/* crit.cpp */
|
854
|
-
|
855
|
-
double calc_kurtosis(size_t ix_arr[], size_t st, size_t end,
|
856
|
-
|
1583
|
+
template <class real_t, class ldouble_safe>
|
1584
|
+
double calc_kurtosis(size_t ix_arr[], size_t st, size_t end, real_t x[], MissingAction missing_action);
|
1585
|
+
template <class real_t, class ldouble_safe>
|
1586
|
+
double calc_kurtosis(real_t x[], size_t n, MissingAction missing_action);
|
1587
|
+
template <class real_t, class mapping, class ldouble_safe>
|
1588
|
+
double calc_kurtosis_weighted(size_t ix_arr[], size_t st, size_t end, real_t x[],
|
1589
|
+
MissingAction missing_action, mapping &restrict w);
|
1590
|
+
template <class real_t, class ldouble_safe>
|
1591
|
+
double calc_kurtosis_weighted(real_t *restrict x, size_t n_, MissingAction missing_action, real_t *restrict w);
|
1592
|
+
template <class real_t, class sparse_ix, class ldouble_safe>
|
1593
|
+
double calc_kurtosis(size_t *restrict ix_arr, size_t st, size_t end, size_t col_num,
|
1594
|
+
real_t Xc[], sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1595
|
+
MissingAction missing_action);
|
1596
|
+
template <class real_t, class sparse_ix, class ldouble_safe>
|
1597
|
+
double calc_kurtosis(size_t col_num, size_t nrows,
|
1598
|
+
real_t Xc[], sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
857
1599
|
MissingAction missing_action);
|
858
|
-
|
1600
|
+
template <class real_t, class sparse_ix, class mapping, class ldouble_safe>
|
1601
|
+
double calc_kurtosis_weighted(size_t *restrict ix_arr, size_t st, size_t end, size_t col_num,
|
1602
|
+
real_t Xc[], sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1603
|
+
MissingAction missing_action, mapping &restrict w);
|
1604
|
+
template <class real_t, class sparse_ix, class ldouble_safe>
|
1605
|
+
double calc_kurtosis_weighted(size_t col_num, size_t nrows,
|
1606
|
+
real_t *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1607
|
+
MissingAction missing_action, real_t *restrict w);
|
1608
|
+
template <class ldouble_safe>
|
1609
|
+
double calc_kurtosis_internal(size_t cnt, int x[], int ncat, size_t buffer_cnt[], double buffer_prob[],
|
1610
|
+
MissingAction missing_action, CategSplit cat_split_type, RNG_engine &rnd_generator);
|
1611
|
+
template <class ldouble_safe>
|
1612
|
+
double calc_kurtosis(size_t *restrict ix_arr, size_t st, size_t end, int x[], int ncat, size_t *restrict buffer_cnt, double buffer_prob[],
|
859
1613
|
MissingAction missing_action, CategSplit cat_split_type, RNG_engine &rnd_generator);
|
860
|
-
|
861
|
-
double
|
862
|
-
|
863
|
-
|
864
|
-
|
865
|
-
|
866
|
-
|
867
|
-
|
868
|
-
|
869
|
-
|
870
|
-
|
871
|
-
|
872
|
-
|
873
|
-
|
874
|
-
|
875
|
-
|
876
|
-
double
|
877
|
-
|
878
|
-
|
1614
|
+
template <class ldouble_safe>
|
1615
|
+
double calc_kurtosis(size_t nrows, int x[], int ncat, size_t buffer_cnt[], double buffer_prob[],
|
1616
|
+
MissingAction missing_action, CategSplit cat_split_type, RNG_engine &rnd_generator);
|
1617
|
+
template <class mapping, class ldouble_safe>
|
1618
|
+
double calc_kurtosis_weighted_internal(std::vector<ldouble_safe> &buffer_cnt, int x[], int ncat,
|
1619
|
+
double buffer_prob[], MissingAction missing_action, CategSplit cat_split_type,
|
1620
|
+
RNG_engine &rnd_generator, mapping &restrict w);
|
1621
|
+
template <class mapping, class ldouble_safe>
|
1622
|
+
double calc_kurtosis_weighted(size_t ix_arr[], size_t st, size_t end, int x[], int ncat, double buffer_prob[],
|
1623
|
+
MissingAction missing_action, CategSplit cat_split_type, RNG_engine &rnd_generator,
|
1624
|
+
mapping &restrict w);
|
1625
|
+
template <class real_t, class ldouble_safe>
|
1626
|
+
double calc_kurtosis_weighted(size_t nrows, int x[], int ncat, double *restrict buffer_prob,
|
1627
|
+
MissingAction missing_action, CategSplit cat_split_type,
|
1628
|
+
RNG_engine &rnd_generator, real_t *restrict w);
|
1629
|
+
template <class int_t, class ldouble_safe>
|
1630
|
+
double expected_sd_cat(double p[], size_t n, int_t pos[]);
|
1631
|
+
template <class number, class int_t, class ldouble_safe>
|
1632
|
+
double expected_sd_cat(number *restrict counts, double *restrict p, size_t n, int_t *restrict pos);
|
1633
|
+
template <class number, class int_t, class ldouble_safe>
|
1634
|
+
double expected_sd_cat_single(number *restrict counts, double *restrict p, size_t n, int_t *restrict pos, size_t cat_exclude, number cnt);
|
1635
|
+
template <class number, class int_t, class ldouble_safe>
|
1636
|
+
double expected_sd_cat_internal(int ncat, number *restrict buffer_cnt, ldouble_safe cnt_l,
|
1637
|
+
int_t *restrict buffer_pos, double *restrict buffer_prob);
|
1638
|
+
template <class int_t, class ldouble_safe>
|
1639
|
+
double expected_sd_cat(size_t *restrict ix_arr, size_t st, size_t end, int x[], int ncat,
|
1640
|
+
MissingAction missing_action,
|
1641
|
+
size_t *restrict buffer_cnt, int_t *restrict buffer_pos, double buffer_prob[]);
|
1642
|
+
template <class mapping, class int_t, class ldouble_safe>
|
1643
|
+
double expected_sd_cat_weighted(size_t *restrict ix_arr, size_t st, size_t end, int x[], int ncat,
|
1644
|
+
MissingAction missing_action, mapping &restrict w,
|
1645
|
+
double *restrict buffer_cnt, int_t *restrict buffer_pos, double *restrict buffer_prob);
|
1646
|
+
template <class number, class ldouble_safe>
|
1647
|
+
double categ_gain(number cnt_left, number cnt_right,
|
1648
|
+
ldouble_safe s_left, ldouble_safe s_right,
|
1649
|
+
ldouble_safe base_info, ldouble_safe cnt);
|
1650
|
+
template <class real_t, class real_t_>
|
1651
|
+
double find_split_rel_gain_t(real_t_ *restrict x, size_t n, double &restrict split_point);
|
1652
|
+
template <class real_t_, class ldouble_safe>
|
1653
|
+
double find_split_rel_gain(real_t_ *restrict x, real_t_ xmean, size_t *restrict ix_arr, size_t st, size_t end, double &restrict split_point, size_t &restrict split_ix);
|
1654
|
+
template <class real_t, class real_t_>
|
1655
|
+
double find_split_rel_gain_t(real_t_ *restrict x, real_t_ xmean, size_t *restrict ix_arr, size_t st, size_t end, double &split_point, size_t &restrict split_ix);
|
1656
|
+
template <class real_t_, class ldouble_safe>
|
1657
|
+
double find_split_rel_gain(real_t_ *restrict x, real_t_ xmean, size_t ix_arr[], size_t st, size_t end, double &split_point, size_t &split_ix);
|
1658
|
+
template <class real_t, class real_t_, class mapping>
|
1659
|
+
double find_split_rel_gain_weighted_t(real_t_ *restrict x, real_t_ xmean, size_t *restrict ix_arr, size_t st, size_t end, double &split_point, size_t &restrict split_ix, mapping &restrict w);
|
1660
|
+
template <class real_t_, class mapping, class ldouble_safe>
|
1661
|
+
double find_split_rel_gain_weighted(real_t_ *restrict x, real_t_ xmean, size_t *restrict ix_arr, size_t st, size_t end, double &restrict split_point, size_t &restrict split_ix, mapping &restrict w);
|
1662
|
+
template <class real_t, class real_t_=double>
|
1663
|
+
real_t calc_sd_right_to_left(real_t_ *restrict x, size_t n, double *restrict sd_arr);
|
1664
|
+
template <class real_t_, class ldouble_safe>
|
1665
|
+
ldouble_safe calc_sd_right_to_left_weighted(real_t_ *restrict x, size_t n, double *restrict sd_arr,
|
1666
|
+
double *restrict w, ldouble_safe &cumw, size_t *restrict sorted_ix);
|
1667
|
+
template <class real_t, class real_t_>
|
1668
|
+
real_t calc_sd_right_to_left(real_t_ *restrict x, real_t_ xmean, size_t ix_arr[], size_t st, size_t end, double *restrict sd_arr);
|
1669
|
+
template <class real_t_, class mapping, class ldouble_safe>
|
1670
|
+
ldouble_safe calc_sd_right_to_left_weighted(real_t_ *restrict x, real_t_ xmean, size_t ix_arr[], size_t st, size_t end,
|
1671
|
+
double *restrict sd_arr, mapping &restrict w, ldouble_safe &cumw);
|
1672
|
+
template <class real_t, class real_t_>
|
1673
|
+
double find_split_std_gain_t(real_t_ *restrict x, size_t n, double *restrict sd_arr,
|
1674
|
+
GainCriterion criterion, double min_gain, double &restrict split_point);
|
1675
|
+
template <class real_t_, class ldouble_safe>
|
1676
|
+
double find_split_std_gain(real_t_ *restrict x, size_t n, double *restrict sd_arr,
|
1677
|
+
GainCriterion criterion, double min_gain, double &restrict split_point);
|
1678
|
+
template <class real_t, class ldouble_safe>
|
1679
|
+
double find_split_std_gain_weighted(real_t *restrict x, size_t n, double *restrict sd_arr,
|
1680
|
+
GainCriterion criterion, double min_gain, double &restrict split_point,
|
1681
|
+
double *restrict w, size_t *restrict sorted_ix);
|
1682
|
+
template <class real_t, class real_t_>
|
1683
|
+
double find_split_std_gain_t(real_t_ *restrict x, real_t_ xmean, size_t ix_arr[], size_t st, size_t end, double *restrict sd_arr,
|
1684
|
+
GainCriterion criterion, double min_gain, double &restrict split_point, size_t &restrict split_ix);
|
1685
|
+
template <class real_t_, class ldouble_safe>
|
1686
|
+
double find_split_std_gain(real_t_ *restrict x, real_t_ xmean, size_t ix_arr[], size_t st, size_t end, double *restrict sd_arr,
|
1687
|
+
GainCriterion criterion, double min_gain, double &restrict split_point, size_t &restrict split_ix);
|
1688
|
+
template <class real_t, class mapping, class ldouble_safe>
|
1689
|
+
double find_split_std_gain_weighted(real_t *restrict x, real_t xmean, size_t ix_arr[], size_t st, size_t end, double *restrict sd_arr,
|
1690
|
+
GainCriterion criterion, double min_gain, double &restrict split_point, size_t &restrict split_ix, mapping &restrict w);
|
1691
|
+
template <class real_t, class ldouble_safe>
|
1692
|
+
double find_split_full_gain(real_t *restrict x, size_t st, size_t end, size_t *restrict ix_arr,
|
1693
|
+
size_t *restrict cols_use, size_t ncols_use, bool force_cols_use,
|
1694
|
+
double *restrict X_row_major, size_t ncols,
|
1695
|
+
double *restrict Xr, size_t *restrict Xr_ind, size_t *restrict Xr_indptr,
|
1696
|
+
double *restrict buffer_sum_left, double *restrict buffer_sum_tot,
|
1697
|
+
size_t &restrict split_ix, double &restrict split_point,
|
1698
|
+
bool x_uses_ix_arr);
|
1699
|
+
template <class real_t, class mapping, class ldouble_safe>
|
1700
|
+
double find_split_full_gain_weighted(real_t *restrict x, size_t st, size_t end, size_t *restrict ix_arr,
|
1701
|
+
size_t *restrict cols_use, size_t ncols_use, bool force_cols_use,
|
1702
|
+
double *restrict X_row_major, size_t ncols,
|
1703
|
+
double *restrict Xr, size_t *restrict Xr_ind, size_t *restrict Xr_indptr,
|
1704
|
+
double *restrict buffer_sum_left, double *restrict buffer_sum_tot,
|
1705
|
+
size_t &restrict split_ix, double &restrict split_point,
|
1706
|
+
bool x_uses_ix_arr,
|
1707
|
+
mapping &restrict w);
|
1708
|
+
template <class real_t_, class real_t>
|
1709
|
+
double find_split_dens_shortform_t(real_t *restrict x, size_t n, double &restrict split_point);
|
1710
|
+
template <class real_t, class ldouble_safe>
|
1711
|
+
double find_split_dens_shortform(real_t *restrict x, size_t n, double &restrict split_point);
|
1712
|
+
template <class real_t_, class real_t, class mapping>
|
1713
|
+
double find_split_dens_shortform_weighted_t(real_t *restrict x, size_t n, double &restrict split_point, mapping &restrict w, size_t *restrict buffer_indices);
|
1714
|
+
template <class real_t, class mapping, class ldouble_safe>
|
1715
|
+
double find_split_dens_shortform_weighted(real_t *restrict x, size_t n, double &restrict split_point, mapping &restrict w, size_t *restrict buffer_indices);
|
1716
|
+
template <class real_t>
|
1717
|
+
double find_split_dens_shortform(real_t *restrict x, size_t *restrict ix_arr, size_t st, size_t end,
|
1718
|
+
double &restrict split_point, size_t &restrict split_ix);
|
1719
|
+
template <class real_t, class mapping>
|
1720
|
+
double find_split_dens_shortform_weighted(real_t *restrict x, size_t *restrict ix_arr, size_t st, size_t end,
|
1721
|
+
double &restrict split_point, size_t &restrict split_ix, mapping &restrict w);
|
1722
|
+
template <class real_t, class ldouble_safe>
|
1723
|
+
double find_split_dens_longform(real_t *restrict x, size_t *restrict ix_arr, size_t st, size_t end,
|
1724
|
+
double &restrict split_point, size_t &restrict split_ix);
|
1725
|
+
template <class real_t, class mapping, class ldouble_safe>
|
1726
|
+
double find_split_dens_longform_weighted(real_t *restrict x, size_t *restrict ix_arr, size_t st, size_t end,
|
1727
|
+
double &restrict split_point, size_t &restrict split_ix, mapping &restrict w);
|
1728
|
+
template <class real_t, class ldouble_safe>
|
1729
|
+
double find_split_dens(real_t *restrict x, size_t *restrict ix_arr, size_t st, size_t end,
|
1730
|
+
double &restrict split_point, size_t &restrict split_ix);
|
1731
|
+
template <class real_t, class mapping, class ldouble_safe>
|
1732
|
+
double find_split_dens_weighted(real_t *restrict x, size_t *restrict ix_arr, size_t st, size_t end,
|
1733
|
+
double &restrict split_point, size_t &restrict split_ix, mapping &restrict w);
|
1734
|
+
template <class int_t, class ldouble_safe>
|
1735
|
+
double find_split_dens_longform(int *restrict x, int ncat, size_t *restrict ix_arr, size_t st, size_t end,
|
1736
|
+
CategSplit cat_split_type, MissingAction missing_action,
|
1737
|
+
int &restrict chosen_cat, signed char *restrict split_categ, int *restrict saved_cat_mode,
|
1738
|
+
size_t *restrict buffer_cnt, int_t *restrict buffer_indices);
|
1739
|
+
template <class mapping, class int_t, class ldouble_safe>
|
1740
|
+
double find_split_dens_longform_weighted(int *restrict x, int ncat, size_t *restrict ix_arr, size_t st, size_t end,
|
1741
|
+
CategSplit cat_split_type, MissingAction missing_action,
|
1742
|
+
int &restrict chosen_cat, signed char *restrict split_categ, int *restrict saved_cat_mode,
|
1743
|
+
int_t *restrict buffer_indices, mapping &restrict w);
|
1744
|
+
template <class ldouble_safe>
|
1745
|
+
double eval_guided_crit(double *restrict x, size_t n, GainCriterion criterion,
|
1746
|
+
double min_gain, bool as_relative_gain, double *restrict buffer_sd,
|
1747
|
+
double &restrict split_point, double &restrict xmin, double &restrict xmax,
|
1748
|
+
size_t *restrict ix_arr_plus_st,
|
1749
|
+
size_t *restrict cols_use, size_t ncols_use, bool force_cols_use,
|
1750
|
+
double *restrict X_row_major, size_t ncols,
|
1751
|
+
double *restrict Xr, size_t *restrict Xr_ind, size_t *restrict Xr_indptr);
|
1752
|
+
template <class ldouble_safe>
|
1753
|
+
double eval_guided_crit_weighted(double *restrict x, size_t n, GainCriterion criterion,
|
1754
|
+
double min_gain, bool as_relative_gain, double *restrict buffer_sd,
|
1755
|
+
double &restrict split_point, double &restrict xmin, double &restrict xmax,
|
1756
|
+
double *restrict w, size_t *restrict buffer_indices,
|
1757
|
+
size_t *restrict ix_arr_plus_st,
|
1758
|
+
size_t *restrict cols_use, size_t ncols_use, bool force_cols_use,
|
1759
|
+
double *restrict X_row_major, size_t ncols,
|
1760
|
+
double *restrict Xr, size_t *restrict Xr_ind, size_t *restrict Xr_indptr);
|
1761
|
+
template <class real_t_, class ldouble_safe>
|
1762
|
+
double eval_guided_crit(size_t *restrict ix_arr, size_t st, size_t end, real_t_ *restrict x,
|
1763
|
+
double *restrict buffer_sd, bool as_relative_gain,
|
1764
|
+
double *restrict buffer_imputed_x, double *restrict saved_xmedian,
|
1765
|
+
size_t &split_ix, double &restrict split_point, double &restrict xmin, double &restrict xmax,
|
1766
|
+
GainCriterion criterion, double min_gain, MissingAction missing_action,
|
1767
|
+
size_t *restrict cols_use, size_t ncols_use, bool force_cols_use,
|
1768
|
+
double *restrict X_row_major, size_t ncols,
|
1769
|
+
double *restrict Xr, size_t *restrict Xr_ind, size_t *restrict Xr_indptr);
|
1770
|
+
template <class real_t_, class mapping, class ldouble_safe>
|
1771
|
+
double eval_guided_crit_weighted(size_t *restrict ix_arr, size_t st, size_t end, real_t_ *restrict x,
|
1772
|
+
double *restrict buffer_sd, bool as_relative_gain,
|
1773
|
+
double *restrict buffer_imputed_x, double *restrict saved_xmedian,
|
1774
|
+
size_t &split_ix, double &restrict split_point, double &restrict xmin, double &restrict xmax,
|
1775
|
+
GainCriterion criterion, double min_gain, MissingAction missing_action,
|
1776
|
+
size_t *restrict cols_use, size_t ncols_use, bool force_cols_use,
|
1777
|
+
double *restrict X_row_major, size_t ncols,
|
1778
|
+
double *restrict Xr, size_t *restrict Xr_ind, size_t *restrict Xr_indptr,
|
1779
|
+
mapping &restrict w);
|
1780
|
+
template <class real_t_, class sparse_ix, class ldouble_safe>
|
879
1781
|
double eval_guided_crit(size_t ix_arr[], size_t st, size_t end,
|
880
|
-
size_t col_num,
|
881
|
-
double buffer_arr[], size_t buffer_pos[],
|
1782
|
+
size_t col_num, real_t_ Xc[], sparse_ix Xc_ind[], sparse_ix Xc_indptr[],
|
1783
|
+
double buffer_arr[], size_t buffer_pos[], bool as_relative_gain,
|
1784
|
+
double *restrict saved_xmedian,
|
882
1785
|
double &split_point, double &xmin, double &xmax,
|
883
|
-
GainCriterion criterion, double min_gain, MissingAction missing_action
|
1786
|
+
GainCriterion criterion, double min_gain, MissingAction missing_action,
|
1787
|
+
size_t *restrict cols_use, size_t ncols_use, bool force_cols_use,
|
1788
|
+
double *restrict X_row_major, size_t ncols,
|
1789
|
+
double *restrict Xr, size_t *restrict Xr_ind, size_t *restrict Xr_indptr);
|
1790
|
+
template <class real_t_, class sparse_ix, class mapping, class ldouble_safe>
|
1791
|
+
double eval_guided_crit_weighted(size_t ix_arr[], size_t st, size_t end,
|
1792
|
+
size_t col_num, real_t_ Xc[], sparse_ix Xc_ind[], sparse_ix Xc_indptr[],
|
1793
|
+
double buffer_arr[], size_t buffer_pos[], bool as_relative_gain,
|
1794
|
+
double *restrict saved_xmedian,
|
1795
|
+
double &restrict split_point, double &restrict xmin, double &restrict xmax,
|
1796
|
+
GainCriterion criterion, double min_gain, MissingAction missing_action,
|
1797
|
+
size_t *restrict cols_use, size_t ncols_use, bool force_cols_use,
|
1798
|
+
double *restrict X_row_major, size_t ncols,
|
1799
|
+
double *restrict Xr, size_t *restrict Xr_ind, size_t *restrict Xr_indptr,
|
1800
|
+
mapping &restrict w);
|
1801
|
+
template <class ldouble_safe>
|
884
1802
|
double eval_guided_crit(size_t *restrict ix_arr, size_t st, size_t end, int *restrict x, int ncat,
|
1803
|
+
int *restrict saved_cat_mode,
|
885
1804
|
size_t *restrict buffer_cnt, size_t *restrict buffer_pos, double *restrict buffer_prob,
|
886
|
-
int &chosen_cat, char *restrict split_categ, char *restrict buffer_split,
|
887
|
-
GainCriterion criterion, double min_gain, bool all_perm,
|
1805
|
+
int &restrict chosen_cat, signed char *restrict split_categ, signed char *restrict buffer_split,
|
1806
|
+
GainCriterion criterion, double min_gain, bool all_perm,
|
1807
|
+
MissingAction missing_action, CategSplit cat_split_type);
|
1808
|
+
template <class mapping, class ldouble_safe>
|
1809
|
+
double eval_guided_crit_weighted(size_t *restrict ix_arr, size_t st, size_t end, int *restrict x, int ncat,
|
1810
|
+
int *restrict saved_cat_mode,
|
1811
|
+
size_t *restrict buffer_pos, double *restrict buffer_prob,
|
1812
|
+
int &restrict chosen_cat, signed char *restrict split_categ, signed char *restrict buffer_split,
|
1813
|
+
GainCriterion criterion, double min_gain, bool all_perm,
|
1814
|
+
MissingAction missing_action, CategSplit cat_split_type,
|
1815
|
+
mapping &restrict w);
|
1816
|
+
|
1817
|
+
/* indexer.cpp */
|
1818
|
+
template <class Tree>
|
1819
|
+
void build_terminal_node_mappings_single_tree(std::vector<size_t> &mappings, size_t &n_terminal, const std::vector<Tree> &tree);
|
1820
|
+
void build_terminal_node_mappings_single_tree(std::vector<size_t> &mappings, size_t &n_terminal, const std::vector<IsoTree> &tree);
|
1821
|
+
void build_terminal_node_mappings_single_tree(std::vector<size_t> &mappings, size_t &n_terminal, const std::vector<IsoHPlane> &tree);
|
1822
|
+
template <class Model>
|
1823
|
+
void build_terminal_node_mappings(TreesIndexer &indexer, const Model &model);
|
1824
|
+
template <class Node>
|
1825
|
+
void build_dindex_recursive
|
1826
|
+
(
|
1827
|
+
const size_t curr_node,
|
1828
|
+
const size_t n_terminal, const size_t ncomb,
|
1829
|
+
const size_t st, const size_t end,
|
1830
|
+
std::vector<size_t> &restrict node_indices, /* array with all terminal indices in 'tree' */
|
1831
|
+
const std::vector<size_t> &restrict node_mappings, /* tree_index : terminal_index */
|
1832
|
+
std::vector<double> &restrict node_distances, /* indexed by terminal_index */
|
1833
|
+
std::vector<double> &restrict node_depths, /* indexed by terminal_index */
|
1834
|
+
size_t curr_depth,
|
1835
|
+
const std::vector<Node> &tree
|
1836
|
+
);
|
1837
|
+
template <class Node>
|
1838
|
+
void build_dindex
|
1839
|
+
(
|
1840
|
+
std::vector<size_t> &restrict node_indices, /* empty, but correctly sized */
|
1841
|
+
const std::vector<size_t> &restrict node_mappings, /* tree_index : terminal_index */
|
1842
|
+
std::vector<double> &restrict node_distances, /* indexed by terminal_index */
|
1843
|
+
std::vector<double> &restrict node_depths, /* indexed by terminal_index */
|
1844
|
+
const size_t n_terminal,
|
1845
|
+
const std::vector<Node> &tree
|
1846
|
+
);
|
1847
|
+
void build_dindex
|
1848
|
+
(
|
1849
|
+
std::vector<size_t> &restrict node_indices, /* empty, but correctly sized */
|
1850
|
+
const std::vector<size_t> &restrict node_mappings, /* tree_index : terminal_index */
|
1851
|
+
std::vector<double> &restrict node_distances, /* indexed by terminal_index */
|
1852
|
+
std::vector<double> &restrict node_depths, /* indexed by terminal_index */
|
1853
|
+
const size_t n_terminal,
|
1854
|
+
const std::vector<IsoTree> &tree
|
1855
|
+
);
|
1856
|
+
void build_dindex
|
1857
|
+
(
|
1858
|
+
std::vector<size_t> &restrict node_indices, /* empty, but correctly sized */
|
1859
|
+
const std::vector<size_t> &restrict node_mappings, /* tree_index : terminal_index */
|
1860
|
+
std::vector<double> &restrict node_distances, /* indexed by terminal_index */
|
1861
|
+
std::vector<double> &restrict node_depths, /* indexed by terminal_index */
|
1862
|
+
const size_t n_terminal,
|
1863
|
+
const std::vector<IsoHPlane> &tree
|
1864
|
+
);
|
1865
|
+
template <class Model>
|
1866
|
+
void build_distance_mappings(TreesIndexer &indexer, const Model &model, int nthreads);
|
1867
|
+
template <class Model>
|
1868
|
+
void build_tree_indices(TreesIndexer &indexer, const Model &model, int nthreads, const bool with_distances);
|
1869
|
+
ISOTREE_EXPORTED
|
1870
|
+
void build_tree_indices(TreesIndexer &indexer, const IsoForest &model, int nthreads, const bool with_distances);
|
1871
|
+
ISOTREE_EXPORTED
|
1872
|
+
void build_tree_indices(TreesIndexer &indexer, const ExtIsoForest &model, int nthreads, const bool with_distances);
|
1873
|
+
ISOTREE_EXPORTED
|
1874
|
+
void build_tree_indices
|
1875
|
+
(
|
1876
|
+
TreesIndexer *indexer,
|
1877
|
+
const IsoForest *model_outputs,
|
1878
|
+
const ExtIsoForest *model_outputs_ext,
|
1879
|
+
int nthreads,
|
1880
|
+
const bool with_distances
|
1881
|
+
);
|
1882
|
+
ISOTREE_EXPORTED
|
1883
|
+
size_t get_number_of_reference_points(const TreesIndexer &indexer) noexcept;
|
1884
|
+
void build_ref_node(SingleTreeIndex &node);
|
1885
|
+
|
1886
|
+
/* ref_indexer.hpp */
|
1887
|
+
template <class Model, class real_t, class sparse_ix>
|
1888
|
+
void set_reference_points(TreesIndexer &indexer, Model &model, const bool with_distances,
|
1889
|
+
real_t *restrict numeric_data, int *restrict categ_data,
|
1890
|
+
bool is_col_major, size_t ld_numeric, size_t ld_categ,
|
1891
|
+
real_t *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1892
|
+
real_t *restrict Xr, sparse_ix *restrict Xr_ind, sparse_ix *restrict Xr_indptr,
|
1893
|
+
size_t nrows, int nthreads);
|
1894
|
+
template <class real_t, class sparse_ix>
|
1895
|
+
void set_reference_points(IsoForest *model_outputs, ExtIsoForest *model_outputs_ext, TreesIndexer *indexer,
|
1896
|
+
const bool with_distances,
|
1897
|
+
real_t *restrict numeric_data, int *restrict categ_data,
|
1898
|
+
bool is_col_major, size_t ld_numeric, size_t ld_categ,
|
1899
|
+
real_t *restrict Xc, sparse_ix *restrict Xc_ind, sparse_ix *restrict Xc_indptr,
|
1900
|
+
real_t *restrict Xr, sparse_ix *restrict Xr_ind, sparse_ix *restrict Xr_indptr,
|
1901
|
+
size_t nrows, int nthreads);
|
888
1902
|
|
889
1903
|
/* merge_models.cpp */
|
1904
|
+
ISOTREE_EXPORTED
|
890
1905
|
void merge_models(IsoForest* model, IsoForest* other,
|
891
1906
|
ExtIsoForest* ext_model, ExtIsoForest* ext_other,
|
892
|
-
Imputer* imputer, Imputer* iother
|
1907
|
+
Imputer* imputer, Imputer* iother,
|
1908
|
+
TreesIndexer* indexer, TreesIndexer* ind_other);
|
1909
|
+
|
1910
|
+
/* subset_models.cpp */
|
1911
|
+
ISOTREE_EXPORTED
|
1912
|
+
void subset_model(IsoForest* model, IsoForest* model_new,
|
1913
|
+
ExtIsoForest* ext_model, ExtIsoForest* ext_model_new,
|
1914
|
+
Imputer* imputer, Imputer* imputer_new,
|
1915
|
+
TreesIndexer* indexer, TreesIndexer* indexer_new,
|
1916
|
+
size_t *trees_take, size_t ntrees_take);
|
893
1917
|
|
894
|
-
#ifdef _ENABLE_CEREAL
|
895
1918
|
/* serialize.cpp */
|
896
|
-
|
897
|
-
void
|
898
|
-
|
899
|
-
void
|
900
|
-
|
901
|
-
void
|
902
|
-
|
903
|
-
|
904
|
-
|
905
|
-
|
906
|
-
|
907
|
-
|
908
|
-
|
909
|
-
|
910
|
-
|
911
|
-
|
912
|
-
|
913
|
-
|
914
|
-
|
915
|
-
|
916
|
-
|
917
|
-
|
918
|
-
|
919
|
-
|
920
|
-
|
921
|
-
|
922
|
-
|
923
|
-
#
|
1919
|
+
[[noreturn]]
|
1920
|
+
void throw_errno();
|
1921
|
+
[[noreturn]]
|
1922
|
+
void throw_ferror(FILE *file);
|
1923
|
+
[[noreturn]]
|
1924
|
+
void throw_feoferr();
|
1925
|
+
class FileHandle
|
1926
|
+
{
|
1927
|
+
public:
|
1928
|
+
FILE *handle = NULL;
|
1929
|
+
FileHandle(const char *fname, const char *mode)
|
1930
|
+
{
|
1931
|
+
this->handle = std::fopen(fname, mode);
|
1932
|
+
if (!(this->handle))
|
1933
|
+
throw_errno();
|
1934
|
+
}
|
1935
|
+
~FileHandle()
|
1936
|
+
{
|
1937
|
+
if (this->handle) {
|
1938
|
+
int err = std::fclose(this->handle);
|
1939
|
+
if (err)
|
1940
|
+
fprintf(stderr, "Error: could not close file.\n");
|
1941
|
+
}
|
1942
|
+
this->handle = NULL;
|
1943
|
+
}
|
1944
|
+
};
|
1945
|
+
|
1946
|
+
#if defined(_WIN32) && (defined(_MSC_VER) || defined(__GNUC__))
|
1947
|
+
#define WCHAR_T_FUNS
|
1948
|
+
#include <stdio.h>
|
1949
|
+
class WFileHandle
|
1950
|
+
{
|
1951
|
+
public:
|
1952
|
+
FILE *handle = NULL;
|
1953
|
+
WFileHandle(const wchar_t *fname, const wchar_t *mode)
|
1954
|
+
{
|
1955
|
+
this->handle = _wfopen(fname, mode);
|
1956
|
+
if (!(this->handle))
|
1957
|
+
throw_errno();
|
1958
|
+
}
|
1959
|
+
~WFileHandle()
|
1960
|
+
{
|
1961
|
+
if (this->handle) {
|
1962
|
+
int err = std::fclose(this->handle);
|
1963
|
+
if (err)
|
1964
|
+
fprintf(stderr, "Error: could not close file.\n");
|
1965
|
+
}
|
1966
|
+
this->handle = NULL;
|
1967
|
+
}
|
1968
|
+
};
|
1969
|
+
#endif
|
1970
|
+
ISOTREE_EXPORTED
|
1971
|
+
bool has_wchar_t_file_serializers() noexcept;
|
1972
|
+
ISOTREE_EXPORTED
|
1973
|
+
size_t determine_serialized_size(const IsoForest &model) noexcept;
|
1974
|
+
ISOTREE_EXPORTED
|
1975
|
+
size_t determine_serialized_size(const ExtIsoForest &model) noexcept;
|
1976
|
+
ISOTREE_EXPORTED
|
1977
|
+
size_t determine_serialized_size(const Imputer &model) noexcept;
|
1978
|
+
ISOTREE_EXPORTED
|
1979
|
+
size_t determine_serialized_size(const TreesIndexer &model) noexcept;
|
1980
|
+
ISOTREE_EXPORTED
|
1981
|
+
void serialize_IsoForest(const IsoForest &model, char *out);
|
1982
|
+
ISOTREE_EXPORTED
|
1983
|
+
void serialize_IsoForest(const IsoForest &model, FILE *out);
|
1984
|
+
ISOTREE_EXPORTED
|
1985
|
+
void serialize_IsoForest(const IsoForest &model, std::ostream &out);
|
1986
|
+
ISOTREE_EXPORTED
|
1987
|
+
std::string serialize_IsoForest(const IsoForest &model);
|
1988
|
+
ISOTREE_EXPORTED
|
1989
|
+
void serialize_IsoForest_ToFile(const IsoForest &model, const char *fname);
|
1990
|
+
#ifdef WCHAR_T_FUNS
|
1991
|
+
ISOTREE_EXPORTED
|
1992
|
+
void serialize_IsoForest_ToFile(const IsoForest &model, const wchar_t *fname);
|
1993
|
+
#endif
|
1994
|
+
ISOTREE_EXPORTED
|
1995
|
+
void deserialize_IsoForest(IsoForest &model, const char *in);
|
1996
|
+
ISOTREE_EXPORTED
|
1997
|
+
void deserialize_IsoForest(IsoForest &model, FILE *in);
|
1998
|
+
ISOTREE_EXPORTED
|
1999
|
+
void deserialize_IsoForest(IsoForest &model, std::istream &in);
|
2000
|
+
ISOTREE_EXPORTED
|
2001
|
+
void deserialize_IsoForest(IsoForest &model, const std::string &in);
|
2002
|
+
ISOTREE_EXPORTED
|
2003
|
+
void deserialize_IsoForest_FromFile(IsoForest &model, const char *fname);
|
2004
|
+
#ifdef WCHAR_T_FUNS
|
2005
|
+
ISOTREE_EXPORTED
|
2006
|
+
void deserialize_IsoForest_FromFile(IsoForest &model, const wchar_t *fname);
|
2007
|
+
#endif
|
2008
|
+
ISOTREE_EXPORTED
|
2009
|
+
void serialize_ExtIsoForest(const ExtIsoForest &model, char *out);
|
2010
|
+
ISOTREE_EXPORTED
|
2011
|
+
void serialize_ExtIsoForest(const ExtIsoForest &model, FILE *out);
|
2012
|
+
ISOTREE_EXPORTED
|
2013
|
+
void serialize_ExtIsoForest(const ExtIsoForest &model, std::ostream &out);
|
2014
|
+
ISOTREE_EXPORTED
|
2015
|
+
std::string serialize_ExtIsoForest(const ExtIsoForest &model);
|
2016
|
+
ISOTREE_EXPORTED
|
2017
|
+
void serialize_ExtIsoForest_ToFile(const ExtIsoForest &model, const char *fname);
|
2018
|
+
#ifdef WCHAR_T_FUNS
|
2019
|
+
ISOTREE_EXPORTED
|
2020
|
+
void serialize_ExtIsoForest_ToFile(const ExtIsoForest &model, const wchar_t *fname);
|
2021
|
+
#endif
|
2022
|
+
ISOTREE_EXPORTED
|
2023
|
+
void deserialize_ExtIsoForest(ExtIsoForest &model, const char *in);
|
2024
|
+
ISOTREE_EXPORTED
|
2025
|
+
void deserialize_ExtIsoForest(ExtIsoForest &model, FILE *in);
|
2026
|
+
ISOTREE_EXPORTED
|
2027
|
+
void deserialize_ExtIsoForest(ExtIsoForest &model, std::istream &in);
|
2028
|
+
ISOTREE_EXPORTED
|
2029
|
+
void deserialize_ExtIsoForest(ExtIsoForest &model, const std::string &in);
|
2030
|
+
ISOTREE_EXPORTED
|
2031
|
+
void deserialize_ExtIsoForest_FromFile(ExtIsoForest &model, const char *fname);
|
2032
|
+
#ifdef WCHAR_T_FUNS
|
2033
|
+
ISOTREE_EXPORTED
|
2034
|
+
void deserialize_ExtIsoForest_FromFile(ExtIsoForest &model, const wchar_t *fname);
|
2035
|
+
#endif
|
2036
|
+
ISOTREE_EXPORTED
|
2037
|
+
void serialize_Imputer(const Imputer &model, char *out);
|
2038
|
+
ISOTREE_EXPORTED
|
2039
|
+
void serialize_Imputer(const Imputer &model, FILE *out);
|
2040
|
+
ISOTREE_EXPORTED
|
2041
|
+
void serialize_Imputer(const Imputer &model, std::ostream &out);
|
2042
|
+
ISOTREE_EXPORTED
|
2043
|
+
std::string serialize_Imputer(const Imputer &model);
|
2044
|
+
ISOTREE_EXPORTED
|
2045
|
+
void serialize_Imputer_ToFile(const Imputer &model, const char *fname);
|
2046
|
+
#ifdef WCHAR_T_FUNS
|
2047
|
+
ISOTREE_EXPORTED
|
2048
|
+
void serialize_Imputer_ToFile(const Imputer &model, const wchar_t *fname);
|
2049
|
+
#endif
|
2050
|
+
ISOTREE_EXPORTED
|
2051
|
+
void deserialize_Imputer(Imputer &model, const char *in);
|
2052
|
+
ISOTREE_EXPORTED
|
2053
|
+
void deserialize_Imputer(Imputer &model, FILE *in);
|
2054
|
+
ISOTREE_EXPORTED
|
2055
|
+
void deserialize_Imputer(Imputer &model, std::istream &in);
|
2056
|
+
ISOTREE_EXPORTED
|
2057
|
+
void deserialize_Imputer(Imputer &model, const std::string &in);
|
2058
|
+
ISOTREE_EXPORTED
|
2059
|
+
void deserialize_Imputer_FromFile(Imputer &model, const char *fname);
|
2060
|
+
#ifdef WCHAR_T_FUNS
|
2061
|
+
ISOTREE_EXPORTED
|
2062
|
+
void deserialize_Imputer_FromFile(Imputer &model, const wchar_t *fname);
|
2063
|
+
#endif
|
2064
|
+
ISOTREE_EXPORTED
|
2065
|
+
void serialize_Indexer(const TreesIndexer &model, char *out);
|
2066
|
+
ISOTREE_EXPORTED
|
2067
|
+
void serialize_Indexer(const TreesIndexer &model, FILE *out);
|
2068
|
+
ISOTREE_EXPORTED
|
2069
|
+
void serialize_Indexer(const TreesIndexer &model, std::ostream &out);
|
2070
|
+
ISOTREE_EXPORTED
|
2071
|
+
std::string serialize_Indexer(const TreesIndexer &model);
|
2072
|
+
ISOTREE_EXPORTED
|
2073
|
+
void serialize_Indexer_ToFile(const TreesIndexer &model, const char *fname);
|
2074
|
+
#ifdef WCHAR_T_FUNS
|
2075
|
+
ISOTREE_EXPORTED
|
2076
|
+
void serialize_Indexer_ToFile(const TreesIndexer &model, const wchar_t *fname);
|
2077
|
+
#endif
|
2078
|
+
ISOTREE_EXPORTED
|
2079
|
+
void deserialize_Indexer(TreesIndexer &model, const char *in);
|
2080
|
+
ISOTREE_EXPORTED
|
2081
|
+
void deserialize_Indexer(TreesIndexer &model, FILE *in);
|
2082
|
+
ISOTREE_EXPORTED
|
2083
|
+
void deserialize_Indexer(TreesIndexer &model, std::istream &in);
|
2084
|
+
ISOTREE_EXPORTED
|
2085
|
+
void deserialize_Indexer(TreesIndexer &model, const std::string &in);
|
2086
|
+
ISOTREE_EXPORTED
|
2087
|
+
void deserialize_Indexer_FromFile(TreesIndexer &model, const char *fname);
|
2088
|
+
#ifdef WCHAR_T_FUNS
|
2089
|
+
ISOTREE_EXPORTED
|
2090
|
+
void deserialize_Indexer_FromFile(TreesIndexer &model, const wchar_t *fname);
|
2091
|
+
#endif
|
2092
|
+
void serialize_isotree(const IsoForest &model, char *out);
|
2093
|
+
void serialize_isotree(const ExtIsoForest &model, char *out);
|
2094
|
+
void serialize_isotree(const Imputer &model, char *out);
|
2095
|
+
void serialize_isotree(const TreesIndexer &model, char *out);
|
2096
|
+
void deserialize_isotree(IsoForest &model, const char *in);
|
2097
|
+
void deserialize_isotree(ExtIsoForest &model, const char *in);
|
2098
|
+
void deserialize_isotree(Imputer &model, const char *in);
|
2099
|
+
void deserialize_isotree(TreesIndexer &model, const char *in);
|
2100
|
+
void incremental_serialize_isotree(const IsoForest &model, char *old_bytes_reallocated);
|
2101
|
+
void incremental_serialize_isotree(const ExtIsoForest &model, char *old_bytes_reallocated);
|
2102
|
+
void incremental_serialize_isotree(const Imputer &model, char *old_bytes_reallocated);
|
2103
|
+
void incremental_serialize_isotree(const TreesIndexer &model, char *old_bytes_reallocated);
|
2104
|
+
ISOTREE_EXPORTED
|
2105
|
+
void incremental_serialize_IsoForest(const IsoForest &model, std::string &old_bytes);
|
2106
|
+
ISOTREE_EXPORTED
|
2107
|
+
void incremental_serialize_ExtIsoForest(const ExtIsoForest &model, std::string &old_bytes);
|
2108
|
+
ISOTREE_EXPORTED
|
2109
|
+
void incremental_serialize_Imputer(const Imputer &model, std::string &old_bytes);
|
2110
|
+
ISOTREE_EXPORTED
|
2111
|
+
void incremental_serialize_Indexer(const TreesIndexer &model, std::string &old_bytes);
|
2112
|
+
ISOTREE_EXPORTED
|
2113
|
+
void inspect_serialized_object
|
2114
|
+
(
|
2115
|
+
const char *serialized_bytes,
|
2116
|
+
bool &is_isotree_model,
|
2117
|
+
bool &is_compatible,
|
2118
|
+
bool &has_combined_objects,
|
2119
|
+
bool &has_IsoForest,
|
2120
|
+
bool &has_ExtIsoForest,
|
2121
|
+
bool &has_Imputer,
|
2122
|
+
bool &has_Indexer,
|
2123
|
+
bool &has_metadata,
|
2124
|
+
size_t &size_metadata
|
2125
|
+
);
|
2126
|
+
ISOTREE_EXPORTED
|
2127
|
+
void inspect_serialized_object
|
2128
|
+
(
|
2129
|
+
FILE *serialized_bytes,
|
2130
|
+
bool &is_isotree_model,
|
2131
|
+
bool &is_compatible,
|
2132
|
+
bool &has_combined_objects,
|
2133
|
+
bool &has_IsoForest,
|
2134
|
+
bool &has_ExtIsoForest,
|
2135
|
+
bool &has_Imputer,
|
2136
|
+
bool &has_Indexer,
|
2137
|
+
bool &has_metadata,
|
2138
|
+
size_t &size_metadata
|
2139
|
+
);
|
2140
|
+
ISOTREE_EXPORTED
|
2141
|
+
void inspect_serialized_object
|
2142
|
+
(
|
2143
|
+
std::istream &serialized_bytes,
|
2144
|
+
bool &is_isotree_model,
|
2145
|
+
bool &is_compatible,
|
2146
|
+
bool &has_combined_objects,
|
2147
|
+
bool &has_IsoForest,
|
2148
|
+
bool &has_ExtIsoForest,
|
2149
|
+
bool &has_Imputer,
|
2150
|
+
bool &has_Indexer,
|
2151
|
+
bool &has_metadata,
|
2152
|
+
size_t &size_metadata
|
2153
|
+
);
|
2154
|
+
ISOTREE_EXPORTED
|
2155
|
+
void inspect_serialized_object
|
2156
|
+
(
|
2157
|
+
const std::string &serialized_bytes,
|
2158
|
+
bool &is_isotree_model,
|
2159
|
+
bool &is_compatible,
|
2160
|
+
bool &has_combined_objects,
|
2161
|
+
bool &has_IsoForest,
|
2162
|
+
bool &has_ExtIsoForest,
|
2163
|
+
bool &has_Imputer,
|
2164
|
+
bool &has_Indexer,
|
2165
|
+
bool &has_metadata,
|
2166
|
+
size_t &size_metadata
|
2167
|
+
);
|
2168
|
+
ISOTREE_EXPORTED
|
2169
|
+
bool check_can_undergo_incremental_serialization(const IsoForest &model, const char *serialized_bytes);
|
2170
|
+
ISOTREE_EXPORTED
|
2171
|
+
bool check_can_undergo_incremental_serialization(const ExtIsoForest &model, const char *serialized_bytes);
|
2172
|
+
ISOTREE_EXPORTED
|
2173
|
+
bool check_can_undergo_incremental_serialization(const Imputer &model, const char *serialized_bytes);
|
2174
|
+
ISOTREE_EXPORTED
|
2175
|
+
bool check_can_undergo_incremental_serialization(const TreesIndexer &model, const char *serialized_bytes);
|
2176
|
+
ISOTREE_EXPORTED
|
2177
|
+
size_t determine_serialized_size_additional_trees(const IsoForest &model, size_t old_ntrees) noexcept;
|
2178
|
+
ISOTREE_EXPORTED
|
2179
|
+
size_t determine_serialized_size_additional_trees(const ExtIsoForest &model, size_t old_ntrees) noexcept;
|
2180
|
+
ISOTREE_EXPORTED
|
2181
|
+
size_t determine_serialized_size_additional_trees(const Imputer &model, size_t old_ntrees) noexcept;
|
2182
|
+
ISOTREE_EXPORTED
|
2183
|
+
size_t determine_serialized_size_additional_trees(const TreesIndexer &model, size_t old_ntrees) noexcept;
|
2184
|
+
ISOTREE_EXPORTED
|
2185
|
+
void incremental_serialize_IsoForest(const IsoForest &model, char *old_bytes_reallocated);
|
2186
|
+
ISOTREE_EXPORTED
|
2187
|
+
void incremental_serialize_ExtIsoForest(const ExtIsoForest &model, char *old_bytes_reallocated);
|
2188
|
+
ISOTREE_EXPORTED
|
2189
|
+
void incremental_serialize_Imputer(const Imputer &model, char *old_bytes_reallocated);
|
2190
|
+
ISOTREE_EXPORTED
|
2191
|
+
void incremental_serialize_Indexer(const TreesIndexer &model, char *old_bytes_reallocated);
|
2192
|
+
ISOTREE_EXPORTED
|
2193
|
+
size_t determine_serialized_size_combined
|
2194
|
+
(
|
2195
|
+
const IsoForest *model,
|
2196
|
+
const ExtIsoForest *model_ext,
|
2197
|
+
const Imputer *imputer,
|
2198
|
+
const TreesIndexer *indexer,
|
2199
|
+
const size_t size_optional_metadata
|
2200
|
+
) noexcept;
|
2201
|
+
ISOTREE_EXPORTED
|
2202
|
+
size_t determine_serialized_size_combined
|
2203
|
+
(
|
2204
|
+
const char *serialized_model,
|
2205
|
+
const char *serialized_model_ext,
|
2206
|
+
const char *serialized_imputer,
|
2207
|
+
const char *serialized_indexer,
|
2208
|
+
const size_t size_optional_metadata
|
2209
|
+
) noexcept;
|
2210
|
+
ISOTREE_EXPORTED
|
2211
|
+
void serialize_combined
|
2212
|
+
(
|
2213
|
+
const IsoForest *model,
|
2214
|
+
const ExtIsoForest *model_ext,
|
2215
|
+
const Imputer *imputer,
|
2216
|
+
const TreesIndexer *indexer,
|
2217
|
+
const char *optional_metadata,
|
2218
|
+
const size_t size_optional_metadata,
|
2219
|
+
char *out
|
2220
|
+
);
|
2221
|
+
ISOTREE_EXPORTED
|
2222
|
+
void serialize_combined
|
2223
|
+
(
|
2224
|
+
const IsoForest *model,
|
2225
|
+
const ExtIsoForest *model_ext,
|
2226
|
+
const Imputer *imputer,
|
2227
|
+
const TreesIndexer *indexer,
|
2228
|
+
const char *optional_metadata,
|
2229
|
+
const size_t size_optional_metadata,
|
2230
|
+
FILE *out
|
2231
|
+
);
|
2232
|
+
ISOTREE_EXPORTED
|
2233
|
+
void serialize_combined
|
2234
|
+
(
|
2235
|
+
const IsoForest *model,
|
2236
|
+
const ExtIsoForest *model_ext,
|
2237
|
+
const Imputer *imputer,
|
2238
|
+
const TreesIndexer *indexer,
|
2239
|
+
const char *optional_metadata,
|
2240
|
+
const size_t size_optional_metadata,
|
2241
|
+
std::ostream &out
|
2242
|
+
);
|
2243
|
+
ISOTREE_EXPORTED
|
2244
|
+
std::string serialize_combined
|
2245
|
+
(
|
2246
|
+
const IsoForest *model,
|
2247
|
+
const ExtIsoForest *model_ext,
|
2248
|
+
const Imputer *imputer,
|
2249
|
+
const TreesIndexer *indexer,
|
2250
|
+
const char *optional_metadata,
|
2251
|
+
const size_t size_optional_metadata
|
2252
|
+
);
|
2253
|
+
ISOTREE_EXPORTED
|
2254
|
+
void serialize_combined
|
2255
|
+
(
|
2256
|
+
const char *serialized_model,
|
2257
|
+
const char *serialized_model_ext,
|
2258
|
+
const char *serialized_imputer,
|
2259
|
+
const char *serialized_indexer,
|
2260
|
+
const char *optional_metadata,
|
2261
|
+
const size_t size_optional_metadata,
|
2262
|
+
FILE *out
|
2263
|
+
);
|
2264
|
+
ISOTREE_EXPORTED
|
2265
|
+
void serialize_combined
|
2266
|
+
(
|
2267
|
+
const char *serialized_model,
|
2268
|
+
const char *serialized_model_ext,
|
2269
|
+
const char *serialized_imputer,
|
2270
|
+
const char *serialized_indexer,
|
2271
|
+
const char *optional_metadata,
|
2272
|
+
const size_t size_optional_metadata,
|
2273
|
+
std::ostream &out
|
2274
|
+
);
|
2275
|
+
ISOTREE_EXPORTED
|
2276
|
+
std::string serialize_combined
|
2277
|
+
(
|
2278
|
+
const char *serialized_model,
|
2279
|
+
const char *serialized_model_ext,
|
2280
|
+
const char *serialized_imputer,
|
2281
|
+
const char *serialized_indexer,
|
2282
|
+
const char *optional_metadata,
|
2283
|
+
const size_t size_optional_metadata
|
2284
|
+
);
|
2285
|
+
ISOTREE_EXPORTED
|
2286
|
+
void deserialize_combined
|
2287
|
+
(
|
2288
|
+
const char* in,
|
2289
|
+
IsoForest *model,
|
2290
|
+
ExtIsoForest *model_ext,
|
2291
|
+
Imputer *imputer,
|
2292
|
+
TreesIndexer *indexer,
|
2293
|
+
char *optional_metadata
|
2294
|
+
);
|
2295
|
+
ISOTREE_EXPORTED
|
2296
|
+
void deserialize_combined
|
2297
|
+
(
|
2298
|
+
FILE* in,
|
2299
|
+
IsoForest *model,
|
2300
|
+
ExtIsoForest *model_ext,
|
2301
|
+
Imputer *imputer,
|
2302
|
+
TreesIndexer *indexer,
|
2303
|
+
char *optional_metadata
|
2304
|
+
);
|
2305
|
+
ISOTREE_EXPORTED
|
2306
|
+
void deserialize_combined
|
2307
|
+
(
|
2308
|
+
std::istream &in,
|
2309
|
+
IsoForest *model,
|
2310
|
+
ExtIsoForest *model_ext,
|
2311
|
+
Imputer *imputer,
|
2312
|
+
TreesIndexer *indexer,
|
2313
|
+
char *optional_metadata
|
2314
|
+
);
|
2315
|
+
ISOTREE_EXPORTED
|
2316
|
+
void deserialize_combined
|
2317
|
+
(
|
2318
|
+
const std::string &in,
|
2319
|
+
IsoForest *model,
|
2320
|
+
ExtIsoForest *model_ext,
|
2321
|
+
Imputer *imputer,
|
2322
|
+
TreesIndexer *indexer,
|
2323
|
+
char *optional_metadata
|
2324
|
+
);
|
2325
|
+
bool check_model_has_range_penalty(const IsoForest &model) noexcept;
|
2326
|
+
bool check_model_has_range_penalty(const ExtIsoForest &model) noexcept;
|
2327
|
+
void add_range_penalty(IsoForest &model) noexcept;
|
2328
|
+
void add_range_penalty(ExtIsoForest &model) noexcept;
|
2329
|
+
void add_range_penalty(Imputer &model) noexcept;
|
2330
|
+
void add_range_penalty(TreesIndexer &model) noexcept;
|
924
2331
|
|
925
2332
|
/* sql.cpp */
|
2333
|
+
ISOTREE_EXPORTED
|
926
2334
|
std::vector<std::string> generate_sql(IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
927
2335
|
std::vector<std::string> &numeric_colnames, std::vector<std::string> &categ_colnames,
|
928
2336
|
std::vector<std::vector<std::string>> &categ_levels,
|
929
2337
|
bool output_tree_num, bool index1, bool single_tree, size_t tree_num,
|
930
2338
|
int nthreads);
|
2339
|
+
ISOTREE_EXPORTED
|
931
2340
|
std::string generate_sql_with_select_from(IsoForest *model_outputs, ExtIsoForest *model_outputs_ext,
|
932
2341
|
std::string &table_from, std::string &select_as,
|
933
2342
|
std::vector<std::string> &numeric_colnames, std::vector<std::string> &categ_colnames,
|
@@ -935,7 +2344,8 @@ std::string generate_sql_with_select_from(IsoForest *model_outputs, ExtIsoForest
|
|
935
2344
|
bool index1, int nthreads);
|
936
2345
|
void generate_tree_rules(std::vector<IsoTree> *trees, std::vector<IsoHPlane> *hplanes, bool output_score,
|
937
2346
|
size_t curr_ix, bool index1, std::string &prev_cond, std::vector<std::string> &node_rules,
|
938
|
-
std::vector<std::string> &conditions_left, std::vector<std::string> &conditions_right
|
2347
|
+
std::vector<std::string> &conditions_left, std::vector<std::string> &conditions_right,
|
2348
|
+
const IsoForest *model_outputs, const ExtIsoForest *model_outputs_ext);
|
939
2349
|
void extract_cond_isotree(IsoForest &model, IsoTree &tree,
|
940
2350
|
std::string &cond_left, std::string &cond_right,
|
941
2351
|
std::vector<std::string> &numeric_colnames, std::vector<std::string> &categ_colnames,
|
@@ -945,7 +2355,9 @@ void extract_cond_ext_isotree(ExtIsoForest &model, IsoHPlane &hplane,
|
|
945
2355
|
std::vector<std::string> &numeric_colnames, std::vector<std::string> &categ_colnames,
|
946
2356
|
std::vector<std::vector<std::string>> &categ_levels);
|
947
2357
|
|
948
|
-
|
949
|
-
|
950
|
-
|
951
|
-
|
2358
|
+
#ifndef _FOR_R
|
2359
|
+
#if defined(__clang__)
|
2360
|
+
#pragma clang diagnostic pop
|
2361
|
+
#endif
|
2362
|
+
#endif
|
2363
|
+
#endif /* ISOTREE_H */
|