datasketches 0.3.1 → 0.4.0

Sign up to get free protection for your applications and to get access to all the features.
Files changed (245) hide show
  1. checksums.yaml +4 -4
  2. data/CHANGELOG.md +10 -0
  3. data/NOTICE +1 -1
  4. data/README.md +0 -2
  5. data/ext/datasketches/cpc_wrapper.cpp +3 -3
  6. data/ext/datasketches/kll_wrapper.cpp +0 -10
  7. data/lib/datasketches/version.rb +1 -1
  8. data/lib/datasketches.rb +1 -1
  9. data/vendor/datasketches-cpp/CMakeLists.txt +23 -20
  10. data/vendor/datasketches-cpp/CODE_OF_CONDUCT.md +3 -0
  11. data/vendor/datasketches-cpp/CONTRIBUTING.md +50 -0
  12. data/vendor/datasketches-cpp/Doxyfile +2827 -0
  13. data/vendor/datasketches-cpp/LICENSE +0 -76
  14. data/vendor/datasketches-cpp/README.md +1 -3
  15. data/vendor/datasketches-cpp/common/CMakeLists.txt +12 -11
  16. data/vendor/datasketches-cpp/common/include/MurmurHash3.h +25 -27
  17. data/vendor/datasketches-cpp/common/include/common_defs.hpp +15 -10
  18. data/vendor/datasketches-cpp/common/include/count_zeros.hpp +11 -2
  19. data/vendor/datasketches-cpp/common/include/kolmogorov_smirnov.hpp +9 -6
  20. data/vendor/datasketches-cpp/common/include/memory_operations.hpp +5 -4
  21. data/vendor/datasketches-cpp/common/include/optional.hpp +148 -0
  22. data/vendor/datasketches-cpp/common/include/quantiles_sorted_view.hpp +95 -2
  23. data/vendor/datasketches-cpp/common/include/quantiles_sorted_view_impl.hpp +1 -1
  24. data/vendor/datasketches-cpp/common/include/serde.hpp +69 -20
  25. data/vendor/datasketches-cpp/common/test/CMakeLists.txt +2 -2
  26. data/vendor/datasketches-cpp/common/test/integration_test.cpp +6 -0
  27. data/vendor/datasketches-cpp/common/test/optional_test.cpp +85 -0
  28. data/vendor/datasketches-cpp/common/test/test_allocator.hpp +14 -14
  29. data/vendor/datasketches-cpp/{python/src/__init__.py → count/CMakeLists.txt} +25 -1
  30. data/vendor/datasketches-cpp/count/include/count_min.hpp +405 -0
  31. data/vendor/datasketches-cpp/count/include/count_min_impl.hpp +497 -0
  32. data/vendor/datasketches-cpp/{MANIFEST.in → count/test/CMakeLists.txt} +23 -20
  33. data/vendor/datasketches-cpp/count/test/count_min_allocation_test.cpp +155 -0
  34. data/vendor/datasketches-cpp/count/test/count_min_test.cpp +303 -0
  35. data/vendor/datasketches-cpp/cpc/include/cpc_common.hpp +14 -20
  36. data/vendor/datasketches-cpp/cpc/include/cpc_compressor.hpp +7 -4
  37. data/vendor/datasketches-cpp/cpc/include/cpc_compressor_impl.hpp +17 -17
  38. data/vendor/datasketches-cpp/cpc/include/cpc_confidence.hpp +3 -3
  39. data/vendor/datasketches-cpp/cpc/include/cpc_sketch.hpp +40 -40
  40. data/vendor/datasketches-cpp/cpc/include/cpc_sketch_impl.hpp +14 -11
  41. data/vendor/datasketches-cpp/cpc/include/cpc_union.hpp +35 -11
  42. data/vendor/datasketches-cpp/cpc/include/cpc_union_impl.hpp +8 -8
  43. data/vendor/datasketches-cpp/cpc/include/cpc_util.hpp +16 -8
  44. data/vendor/datasketches-cpp/cpc/include/u32_table.hpp +3 -2
  45. data/vendor/datasketches-cpp/cpc/include/u32_table_impl.hpp +5 -5
  46. data/vendor/datasketches-cpp/cpc/test/CMakeLists.txt +20 -7
  47. data/vendor/datasketches-cpp/cpc/test/cpc_sketch_deserialize_from_java_test.cpp +60 -0
  48. data/vendor/datasketches-cpp/cpc/test/cpc_sketch_serialize_for_java.cpp +38 -0
  49. data/vendor/datasketches-cpp/cpc/test/cpc_sketch_test.cpp +4 -29
  50. data/vendor/datasketches-cpp/cpc/test/cpc_union_test.cpp +4 -4
  51. data/vendor/datasketches-cpp/{tox.ini → density/CMakeLists.txt} +24 -8
  52. data/vendor/datasketches-cpp/density/include/density_sketch.hpp +256 -0
  53. data/vendor/datasketches-cpp/density/include/density_sketch_impl.hpp +543 -0
  54. data/vendor/datasketches-cpp/{python/datasketches/__init__.py → density/test/CMakeLists.txt} +15 -3
  55. data/vendor/datasketches-cpp/density/test/density_sketch_test.cpp +244 -0
  56. data/vendor/datasketches-cpp/fi/include/frequent_items_sketch.hpp +21 -9
  57. data/vendor/datasketches-cpp/fi/include/frequent_items_sketch_impl.hpp +6 -4
  58. data/vendor/datasketches-cpp/fi/include/reverse_purge_hash_map.hpp +9 -3
  59. data/vendor/datasketches-cpp/fi/test/CMakeLists.txt +14 -1
  60. data/vendor/datasketches-cpp/fi/test/frequent_items_sketch_deserialize_from_java_test.cpp +95 -0
  61. data/vendor/datasketches-cpp/fi/test/frequent_items_sketch_serialize_for_java.cpp +83 -0
  62. data/vendor/datasketches-cpp/fi/test/frequent_items_sketch_test.cpp +3 -42
  63. data/vendor/datasketches-cpp/hll/include/CouponList-internal.hpp +2 -2
  64. data/vendor/datasketches-cpp/hll/include/CouponList.hpp +3 -1
  65. data/vendor/datasketches-cpp/hll/include/Hll4Array-internal.hpp +19 -11
  66. data/vendor/datasketches-cpp/hll/include/Hll4Array.hpp +2 -5
  67. data/vendor/datasketches-cpp/hll/include/Hll6Array-internal.hpp +19 -7
  68. data/vendor/datasketches-cpp/hll/include/Hll6Array.hpp +1 -1
  69. data/vendor/datasketches-cpp/hll/include/Hll8Array-internal.hpp +98 -42
  70. data/vendor/datasketches-cpp/hll/include/Hll8Array.hpp +2 -0
  71. data/vendor/datasketches-cpp/hll/include/HllArray-internal.hpp +94 -61
  72. data/vendor/datasketches-cpp/hll/include/HllArray.hpp +20 -8
  73. data/vendor/datasketches-cpp/hll/include/HllSketch-internal.hpp +4 -4
  74. data/vendor/datasketches-cpp/hll/include/HllSketchImpl.hpp +3 -1
  75. data/vendor/datasketches-cpp/hll/include/HllSketchImplFactory.hpp +3 -21
  76. data/vendor/datasketches-cpp/hll/include/HllUnion-internal.hpp +8 -0
  77. data/vendor/datasketches-cpp/hll/include/HllUtil.hpp +14 -18
  78. data/vendor/datasketches-cpp/hll/include/coupon_iterator-internal.hpp +1 -1
  79. data/vendor/datasketches-cpp/hll/include/coupon_iterator.hpp +8 -2
  80. data/vendor/datasketches-cpp/hll/include/hll.hpp +79 -65
  81. data/vendor/datasketches-cpp/hll/test/CMakeLists.txt +14 -1
  82. data/vendor/datasketches-cpp/hll/test/HllUnionTest.cpp +7 -1
  83. data/vendor/datasketches-cpp/hll/test/ToFromByteArrayTest.cpp +0 -68
  84. data/vendor/datasketches-cpp/hll/test/hll_sketch_deserialize_from_java_test.cpp +69 -0
  85. data/vendor/datasketches-cpp/hll/test/hll_sketch_serialize_for_java.cpp +52 -0
  86. data/vendor/datasketches-cpp/kll/include/kll_helper.hpp +0 -1
  87. data/vendor/datasketches-cpp/kll/include/kll_helper_impl.hpp +2 -2
  88. data/vendor/datasketches-cpp/kll/include/kll_sketch.hpp +79 -53
  89. data/vendor/datasketches-cpp/kll/include/kll_sketch_impl.hpp +61 -132
  90. data/vendor/datasketches-cpp/kll/test/CMakeLists.txt +14 -1
  91. data/vendor/datasketches-cpp/kll/test/kll_sketch_deserialize_from_java_test.cpp +103 -0
  92. data/vendor/datasketches-cpp/kll/test/kll_sketch_serialize_for_java.cpp +62 -0
  93. data/vendor/datasketches-cpp/kll/test/kll_sketch_test.cpp +5 -40
  94. data/vendor/datasketches-cpp/quantiles/include/quantiles_sketch.hpp +76 -54
  95. data/vendor/datasketches-cpp/quantiles/include/quantiles_sketch_impl.hpp +66 -136
  96. data/vendor/datasketches-cpp/quantiles/test/CMakeLists.txt +14 -1
  97. data/vendor/datasketches-cpp/quantiles/test/quantiles_sketch_deserialize_from_java_test.cpp +84 -0
  98. data/vendor/datasketches-cpp/quantiles/test/quantiles_sketch_serialize_for_java.cpp +52 -0
  99. data/vendor/datasketches-cpp/quantiles/test/quantiles_sketch_test.cpp +15 -39
  100. data/vendor/datasketches-cpp/req/include/req_common.hpp +7 -3
  101. data/vendor/datasketches-cpp/req/include/req_compactor_impl.hpp +2 -4
  102. data/vendor/datasketches-cpp/req/include/req_sketch.hpp +105 -26
  103. data/vendor/datasketches-cpp/req/include/req_sketch_impl.hpp +50 -111
  104. data/vendor/datasketches-cpp/req/test/CMakeLists.txt +14 -1
  105. data/vendor/datasketches-cpp/req/test/req_sketch_deserialize_from_java_test.cpp +55 -0
  106. data/vendor/datasketches-cpp/{tuple/include/array_of_doubles_intersection_impl.hpp → req/test/req_sketch_serialize_for_java.cpp} +12 -7
  107. data/vendor/datasketches-cpp/req/test/req_sketch_test.cpp +3 -89
  108. data/vendor/datasketches-cpp/sampling/CMakeLists.txt +4 -0
  109. data/vendor/datasketches-cpp/sampling/include/ebpps_sample.hpp +210 -0
  110. data/vendor/datasketches-cpp/sampling/include/ebpps_sample_impl.hpp +535 -0
  111. data/vendor/datasketches-cpp/sampling/include/ebpps_sketch.hpp +281 -0
  112. data/vendor/datasketches-cpp/sampling/include/ebpps_sketch_impl.hpp +531 -0
  113. data/vendor/datasketches-cpp/sampling/include/var_opt_sketch.hpp +89 -32
  114. data/vendor/datasketches-cpp/sampling/include/var_opt_sketch_impl.hpp +33 -19
  115. data/vendor/datasketches-cpp/sampling/include/var_opt_union.hpp +13 -10
  116. data/vendor/datasketches-cpp/sampling/include/var_opt_union_impl.hpp +23 -19
  117. data/vendor/datasketches-cpp/sampling/test/CMakeLists.txt +55 -8
  118. data/vendor/datasketches-cpp/sampling/test/ebpps_allocation_test.cpp +96 -0
  119. data/vendor/datasketches-cpp/sampling/test/ebpps_sample_test.cpp +137 -0
  120. data/vendor/datasketches-cpp/sampling/test/ebpps_sketch_test.cpp +266 -0
  121. data/vendor/datasketches-cpp/sampling/test/var_opt_sketch_deserialize_from_java_test.cpp +81 -0
  122. data/vendor/datasketches-cpp/sampling/test/var_opt_sketch_serialize_for_java.cpp +54 -0
  123. data/vendor/datasketches-cpp/sampling/test/var_opt_sketch_test.cpp +33 -51
  124. data/vendor/datasketches-cpp/sampling/test/var_opt_union_deserialize_from_java_test.cpp +50 -0
  125. data/vendor/datasketches-cpp/sampling/test/var_opt_union_serialize_for_java.cpp +56 -0
  126. data/vendor/datasketches-cpp/sampling/test/var_opt_union_test.cpp +0 -20
  127. data/vendor/datasketches-cpp/theta/CMakeLists.txt +1 -0
  128. data/vendor/datasketches-cpp/theta/include/bit_packing.hpp +6279 -0
  129. data/vendor/datasketches-cpp/theta/include/bounds_on_ratios_in_sampled_sets.hpp +1 -0
  130. data/vendor/datasketches-cpp/theta/include/bounds_on_ratios_in_theta_sketched_sets.hpp +7 -6
  131. data/vendor/datasketches-cpp/theta/include/compact_theta_sketch_parser.hpp +14 -8
  132. data/vendor/datasketches-cpp/theta/include/compact_theta_sketch_parser_impl.hpp +60 -46
  133. data/vendor/datasketches-cpp/theta/include/theta_a_not_b.hpp +20 -5
  134. data/vendor/datasketches-cpp/theta/include/theta_constants.hpp +10 -4
  135. data/vendor/datasketches-cpp/theta/include/theta_helpers.hpp +3 -1
  136. data/vendor/datasketches-cpp/theta/include/theta_intersection.hpp +13 -5
  137. data/vendor/datasketches-cpp/theta/include/theta_intersection_base_impl.hpp +5 -5
  138. data/vendor/datasketches-cpp/theta/include/theta_intersection_impl.hpp +3 -3
  139. data/vendor/datasketches-cpp/theta/include/theta_jaccard_similarity.hpp +2 -1
  140. data/vendor/datasketches-cpp/theta/include/theta_jaccard_similarity_base.hpp +1 -0
  141. data/vendor/datasketches-cpp/theta/include/theta_set_difference_base_impl.hpp +1 -1
  142. data/vendor/datasketches-cpp/theta/include/theta_sketch.hpp +180 -33
  143. data/vendor/datasketches-cpp/theta/include/theta_sketch_impl.hpp +430 -130
  144. data/vendor/datasketches-cpp/theta/include/theta_union.hpp +17 -10
  145. data/vendor/datasketches-cpp/theta/include/theta_union_base_impl.hpp +10 -10
  146. data/vendor/datasketches-cpp/theta/include/theta_union_impl.hpp +3 -3
  147. data/vendor/datasketches-cpp/theta/include/theta_update_sketch_base.hpp +21 -6
  148. data/vendor/datasketches-cpp/theta/include/theta_update_sketch_base_impl.hpp +13 -3
  149. data/vendor/datasketches-cpp/theta/test/CMakeLists.txt +15 -1
  150. data/vendor/datasketches-cpp/theta/test/bit_packing_test.cpp +80 -0
  151. data/vendor/datasketches-cpp/theta/test/theta_sketch_deserialize_from_java_test.cpp +57 -0
  152. data/vendor/datasketches-cpp/theta/test/theta_sketch_serialize_for_java.cpp +61 -0
  153. data/vendor/datasketches-cpp/theta/test/theta_sketch_test.cpp +39 -188
  154. data/vendor/datasketches-cpp/theta/test/theta_union_test.cpp +25 -0
  155. data/vendor/datasketches-cpp/tuple/CMakeLists.txt +8 -7
  156. data/vendor/datasketches-cpp/tuple/include/array_of_doubles_sketch.hpp +19 -144
  157. data/vendor/datasketches-cpp/tuple/include/{array_of_doubles_a_not_b.hpp → array_tuple_a_not_b.hpp} +24 -16
  158. data/vendor/datasketches-cpp/tuple/include/{array_of_doubles_a_not_b_impl.hpp → array_tuple_a_not_b_impl.hpp} +4 -4
  159. data/vendor/datasketches-cpp/tuple/include/array_tuple_intersection.hpp +65 -0
  160. data/vendor/datasketches-cpp/tuple/include/array_tuple_intersection_impl.hpp +31 -0
  161. data/vendor/datasketches-cpp/tuple/include/array_tuple_sketch.hpp +237 -0
  162. data/vendor/datasketches-cpp/tuple/include/{array_of_doubles_sketch_impl.hpp → array_tuple_sketch_impl.hpp} +40 -41
  163. data/vendor/datasketches-cpp/tuple/include/array_tuple_union.hpp +81 -0
  164. data/vendor/datasketches-cpp/tuple/include/array_tuple_union_impl.hpp +43 -0
  165. data/vendor/datasketches-cpp/tuple/include/tuple_a_not_b.hpp +11 -2
  166. data/vendor/datasketches-cpp/tuple/include/tuple_intersection.hpp +17 -10
  167. data/vendor/datasketches-cpp/tuple/include/tuple_jaccard_similarity.hpp +2 -1
  168. data/vendor/datasketches-cpp/tuple/include/tuple_sketch.hpp +95 -32
  169. data/vendor/datasketches-cpp/tuple/include/tuple_sketch_impl.hpp +2 -1
  170. data/vendor/datasketches-cpp/tuple/include/tuple_union.hpp +19 -11
  171. data/vendor/datasketches-cpp/tuple/test/CMakeLists.txt +16 -1
  172. data/vendor/datasketches-cpp/tuple/test/aod_sketch_deserialize_from_java_test.cpp +76 -0
  173. data/vendor/datasketches-cpp/tuple/test/aod_sketch_serialize_for_java.cpp +62 -0
  174. data/vendor/datasketches-cpp/tuple/test/array_of_doubles_sketch_test.cpp +5 -129
  175. data/vendor/datasketches-cpp/tuple/test/engagement_test.cpp +85 -89
  176. data/vendor/datasketches-cpp/tuple/test/tuple_jaccard_similarity_test.cpp +3 -1
  177. data/vendor/datasketches-cpp/tuple/test/tuple_sketch_deserialize_from_java_test.cpp +47 -0
  178. data/vendor/datasketches-cpp/{python/src/datasketches.cpp → tuple/test/tuple_sketch_serialize_for_java.cpp} +16 -30
  179. data/vendor/datasketches-cpp/tuple/test/tuple_sketch_test.cpp +1 -1
  180. data/vendor/datasketches-cpp/version.cfg.in +1 -1
  181. metadata +61 -79
  182. data/vendor/datasketches-cpp/fi/test/items_sketch_string_from_java.sk +0 -0
  183. data/vendor/datasketches-cpp/fi/test/items_sketch_string_utf8_from_java.sk +0 -0
  184. data/vendor/datasketches-cpp/fi/test/longs_sketch_from_java.sk +0 -0
  185. data/vendor/datasketches-cpp/hll/test/array6_from_java.sk +0 -0
  186. data/vendor/datasketches-cpp/hll/test/compact_array4_from_java.sk +0 -0
  187. data/vendor/datasketches-cpp/hll/test/compact_set_from_java.sk +0 -0
  188. data/vendor/datasketches-cpp/hll/test/list_from_java.sk +0 -0
  189. data/vendor/datasketches-cpp/hll/test/updatable_array4_from_java.sk +0 -0
  190. data/vendor/datasketches-cpp/hll/test/updatable_set_from_java.sk +0 -0
  191. data/vendor/datasketches-cpp/kll/test/kll_sketch_from_java.sk +0 -0
  192. data/vendor/datasketches-cpp/pyproject.toml +0 -23
  193. data/vendor/datasketches-cpp/python/CMakeLists.txt +0 -81
  194. data/vendor/datasketches-cpp/python/README.md +0 -85
  195. data/vendor/datasketches-cpp/python/datasketches/PySerDe.py +0 -104
  196. data/vendor/datasketches-cpp/python/include/py_serde.hpp +0 -113
  197. data/vendor/datasketches-cpp/python/jupyter/CPCSketch.ipynb +0 -345
  198. data/vendor/datasketches-cpp/python/jupyter/FrequentItemsSketch.ipynb +0 -354
  199. data/vendor/datasketches-cpp/python/jupyter/HLLSketch.ipynb +0 -346
  200. data/vendor/datasketches-cpp/python/jupyter/KLLSketch.ipynb +0 -463
  201. data/vendor/datasketches-cpp/python/jupyter/ThetaSketchNotebook.ipynb +0 -403
  202. data/vendor/datasketches-cpp/python/pybind11Path.cmd +0 -21
  203. data/vendor/datasketches-cpp/python/src/cpc_wrapper.cpp +0 -90
  204. data/vendor/datasketches-cpp/python/src/fi_wrapper.cpp +0 -128
  205. data/vendor/datasketches-cpp/python/src/hll_wrapper.cpp +0 -134
  206. data/vendor/datasketches-cpp/python/src/kll_wrapper.cpp +0 -210
  207. data/vendor/datasketches-cpp/python/src/ks_wrapper.cpp +0 -68
  208. data/vendor/datasketches-cpp/python/src/py_serde.cpp +0 -111
  209. data/vendor/datasketches-cpp/python/src/quantiles_wrapper.cpp +0 -204
  210. data/vendor/datasketches-cpp/python/src/req_wrapper.cpp +0 -215
  211. data/vendor/datasketches-cpp/python/src/theta_wrapper.cpp +0 -172
  212. data/vendor/datasketches-cpp/python/src/vector_of_kll.cpp +0 -490
  213. data/vendor/datasketches-cpp/python/src/vo_wrapper.cpp +0 -173
  214. data/vendor/datasketches-cpp/python/tests/__init__.py +0 -16
  215. data/vendor/datasketches-cpp/python/tests/cpc_test.py +0 -64
  216. data/vendor/datasketches-cpp/python/tests/fi_test.py +0 -110
  217. data/vendor/datasketches-cpp/python/tests/hll_test.py +0 -130
  218. data/vendor/datasketches-cpp/python/tests/kll_test.py +0 -125
  219. data/vendor/datasketches-cpp/python/tests/quantiles_test.py +0 -126
  220. data/vendor/datasketches-cpp/python/tests/req_test.py +0 -126
  221. data/vendor/datasketches-cpp/python/tests/theta_test.py +0 -146
  222. data/vendor/datasketches-cpp/python/tests/vector_of_kll_test.py +0 -148
  223. data/vendor/datasketches-cpp/python/tests/vo_test.py +0 -125
  224. data/vendor/datasketches-cpp/req/test/req_float_empty_from_java.sk +0 -0
  225. data/vendor/datasketches-cpp/req/test/req_float_estimation_from_java.sk +0 -0
  226. data/vendor/datasketches-cpp/req/test/req_float_exact_from_java.sk +0 -0
  227. data/vendor/datasketches-cpp/req/test/req_float_raw_items_from_java.sk +0 -0
  228. data/vendor/datasketches-cpp/req/test/req_float_single_item_from_java.sk +0 -0
  229. data/vendor/datasketches-cpp/sampling/test/binaries_from_java.txt +0 -67
  230. data/vendor/datasketches-cpp/sampling/test/varopt_sketch_long_sampling.sk +0 -0
  231. data/vendor/datasketches-cpp/sampling/test/varopt_sketch_string_exact.sk +0 -0
  232. data/vendor/datasketches-cpp/sampling/test/varopt_union_double_sampling.sk +0 -0
  233. data/vendor/datasketches-cpp/setup.py +0 -110
  234. data/vendor/datasketches-cpp/theta/test/theta_compact_empty_from_java.sk +0 -0
  235. data/vendor/datasketches-cpp/theta/test/theta_compact_estimation_from_java.sk +0 -0
  236. data/vendor/datasketches-cpp/theta/test/theta_compact_exact_from_java.sk +0 -0
  237. data/vendor/datasketches-cpp/theta/test/theta_compact_single_item_from_java.sk +0 -0
  238. data/vendor/datasketches-cpp/tuple/include/array_of_doubles_intersection.hpp +0 -52
  239. data/vendor/datasketches-cpp/tuple/include/array_of_doubles_union.hpp +0 -81
  240. data/vendor/datasketches-cpp/tuple/include/array_of_doubles_union_impl.hpp +0 -43
  241. data/vendor/datasketches-cpp/tuple/test/aod_1_compact_empty_from_java.sk +0 -1
  242. data/vendor/datasketches-cpp/tuple/test/aod_1_compact_estimation_from_java.sk +0 -0
  243. data/vendor/datasketches-cpp/tuple/test/aod_1_compact_non_empty_no_entries_from_java.sk +0 -0
  244. data/vendor/datasketches-cpp/tuple/test/aod_2_compact_exact_from_java.sk +0 -0
  245. data/vendor/datasketches-cpp/tuple/test/aod_3_compact_empty_from_java.sk +0 -1
@@ -0,0 +1,54 @@
1
+ /*
2
+ * Licensed to the Apache Software Foundation (ASF) under one
3
+ * or more contributor license agreements. See the NOTICE file
4
+ * distributed with this work for additional information
5
+ * regarding copyright ownership. The ASF licenses this file
6
+ * to you under the Apache License, Version 2.0 (the
7
+ * "License"); you may not use this file except in compliance
8
+ * with the License. You may obtain a copy of the License at
9
+ *
10
+ * http://www.apache.org/licenses/LICENSE-2.0
11
+ *
12
+ * Unless required by applicable law or agreed to in writing,
13
+ * software distributed under the License is distributed on an
14
+ * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
15
+ * KIND, either express or implied. See the License for the
16
+ * specific language governing permissions and limitations
17
+ * under the License.
18
+ */
19
+
20
+ #include <catch2/catch.hpp>
21
+ #include <fstream>
22
+ #include <var_opt_sketch.hpp>
23
+
24
+ namespace datasketches {
25
+
26
+ TEST_CASE("varopt sketch long generate", "[serialize_for_java]") {
27
+ const unsigned n_arr[] = {0, 1, 10, 100, 1000, 10000, 100000, 1000000};
28
+ for (const unsigned n: n_arr) {
29
+ var_opt_sketch<long> sketch(32);
30
+ for (unsigned i = 1; i <= n; ++i) sketch.update(i);
31
+ std::ofstream os("varopt_sketch_long_n" + std::to_string(n) + "_cpp.sk", std::ios::binary);
32
+ sketch.serialize(os);
33
+ }
34
+ }
35
+
36
+ TEST_CASE("varopt sketch string exact", "[serialize_for_java]") {
37
+ var_opt_sketch<std::string> sketch(1024);
38
+ for (unsigned i = 1; i <= 200; ++i) sketch.update(std::to_string(i), 1000.0 / i);
39
+ std::ofstream os("varopt_sketch_string_exact_cpp.sk", std::ios::binary);
40
+ sketch.serialize(os);
41
+ }
42
+
43
+ TEST_CASE("varopt sketch long sampling", "[serialize_for_java]") {
44
+ var_opt_sketch<long> sketch(1024);
45
+ for (unsigned i = 0; i < 2000; ++i) sketch.update(i);
46
+ // negative heavy items to allow a simple predicate to filter
47
+ sketch.update(-1L, 100000.0);
48
+ sketch.update(-2L, 110000.0);
49
+ sketch.update(-3L, 120000.0);
50
+ std::ofstream os("varopt_sketch_long_sampling_cpp.sk", std::ios::binary);
51
+ sketch.serialize(os);
52
+ }
53
+
54
+ } /* namespace datasketches */
@@ -52,17 +52,15 @@ static void check_if_equal(var_opt_sketch<T, A>& sk1, var_opt_sketch<T, A>& sk2)
52
52
  REQUIRE(sk1.get_k() == sk2.get_k());
53
53
  REQUIRE(sk1.get_n() == sk2.get_n());
54
54
  REQUIRE(sk1.get_num_samples() == sk2.get_num_samples());
55
-
55
+
56
56
  auto it1 = sk1.begin();
57
57
  auto it2 = sk2.begin();
58
- size_t i = 0;
59
58
 
60
59
  while ((it1 != sk1.end()) && (it2 != sk2.end())) {
61
- const std::pair<const T&, const double> p1 = *it1;
62
- const std::pair<const T&, const double> p2 = *it2;
60
+ auto p1 = *it1;
61
+ auto p2 = *it2;
63
62
  REQUIRE(p1.first == p2.first); // data values
64
63
  REQUIRE(p1.second == p2.second); // weights
65
- ++i;
66
64
  ++it1;
67
65
  ++it2;
68
66
  }
@@ -182,7 +180,7 @@ TEST_CASE("varopt sketch: invalid weight", "[var_opt_sketch]") {
182
180
  var_opt_sketch<std::string> sk(100, resize_factor::X2);
183
181
  REQUIRE_THROWS_AS(sk.update("invalid_weight", -1.0), std::invalid_argument);
184
182
 
185
- // should not throw but sketch shoulds till be empty
183
+ // should not throw but sketch should still be empty
186
184
  sk.update("zero weight", 0.0);
187
185
  REQUIRE(sk.is_empty());
188
186
  }
@@ -213,7 +211,7 @@ TEST_CASE("varopt sketch: cumulative weight", "[var_opt_sketch]") {
213
211
 
214
212
  double input_sum = 0.0;
215
213
  for (size_t i = 0; i < n; ++i) {
216
- // generate weights aboev and below 1.0 using w ~ exp(5*N(0,1))
214
+ // generate weights above and below 1.0 using w ~ exp(5*N(0,1))
217
215
  // which covers about 10 orders of magnitude
218
216
  double w = std::exp(5 * N(rand));
219
217
  input_sum += w;
@@ -221,12 +219,12 @@ TEST_CASE("varopt sketch: cumulative weight", "[var_opt_sketch]") {
221
219
  }
222
220
 
223
221
  double output_sum = 0.0;
224
- for (auto it : sk) { // std::pair<int, weight>
225
- output_sum += it.second;
222
+ for (auto pair : sk) { // std::pair<int, weight>
223
+ output_sum += pair.second;
226
224
  }
227
225
 
228
226
  double weight_ratio = output_sum / input_sum;
229
- REQUIRE(std::abs(weight_ratio - 1.0) == Approx(0).margin(EPS));
227
+ REQUIRE(weight_ratio == Approx(1.0).margin(EPS));
230
228
  }
231
229
 
232
230
  TEST_CASE("varopt sketch: under-full sketch serialization", "[var_opt_sketch]") {
@@ -275,26 +273,38 @@ TEST_CASE("varopt sketch: full sketch serialization", "[var_opt_sketch]") {
275
273
  sk.update(100, 100.0);
276
274
  sk.update(101, 101.0);
277
275
 
276
+ subset_summary summary = sk.estimate_subset_sum([](int){ return true; });
277
+ double total_weight = summary.total_sketch_weight;
278
+ double cum_weight = 0.0;
279
+ for (auto pair : sk) {
280
+ cum_weight += pair.second;
281
+ }
282
+ double weight_ratio = cum_weight / total_weight;
283
+ REQUIRE(weight_ratio == Approx(1.0).margin(EPS));
284
+
278
285
  // first 2 entries should be heavy and in heap order (smallest at root)
279
286
  auto it = sk.begin();
280
- const std::pair<const int, const double> p1 = *it;
287
+ auto p1 = *it;
281
288
  ++it;
282
- const std::pair<const int, const double> p2 = *it;
289
+ auto p2 = *it;
283
290
  REQUIRE(p1.second == Approx(100.0).margin(EPS));
284
291
  REQUIRE(p2.second == Approx(101.0).margin(EPS));
285
292
  REQUIRE(p1.first == 100);
286
293
  REQUIRE(p2.first == 101);
294
+ // using operator ->
295
+ REQUIRE(it->first == p2.first);
296
+ REQUIRE(it->second == p2.second);
287
297
 
288
298
  // check for 4 preamble longs
289
299
  auto bytes = sk.serialize();
290
300
  REQUIRE((bytes.data()[0] & 0x3f) == 4);; // PREAMBLE_LONGS_WARMUP
291
301
 
292
- var_opt_sketch<int> sk_from_bytes = var_opt_sketch<int>::deserialize(bytes.data(), bytes.size());
302
+ auto sk_from_bytes = var_opt_sketch<int>::deserialize(bytes.data(), bytes.size());
293
303
  check_if_equal(sk, sk_from_bytes);
294
304
 
295
305
  std::stringstream ss(std::ios::in | std::ios::out | std::ios::binary);
296
306
  sk.serialize(ss);
297
- var_opt_sketch<int> sk_from_stream = var_opt_sketch<int>::deserialize(ss);
307
+ auto sk_from_stream = var_opt_sketch<int>::deserialize(ss);
298
308
  check_if_equal(sk, sk_from_stream);
299
309
 
300
310
  // ensure we unroll properly
@@ -340,6 +350,15 @@ TEST_CASE("varopt sketch: pseudo-light update", "[var_opt_sketch]") {
340
350
  auto it = sk.begin();
341
351
  double wt = (*it).second;
342
352
  REQUIRE(wt == Approx((k + 2.0) / k).margin(EPS));
353
+
354
+ subset_summary summary = sk.estimate_subset_sum([](int){ return true; });
355
+ double total_weight = summary.total_sketch_weight;
356
+ double cum_weight = 0.0;
357
+ for (auto pair : sk) {
358
+ cum_weight += pair.second;
359
+ }
360
+ double weight_ratio = cum_weight / total_weight;
361
+ REQUIRE(weight_ratio == Approx(1.0).margin(EPS));
343
362
  }
344
363
 
345
364
  TEST_CASE("varopt sketch: pseudo-heavy update", "[var_opt_sketch]") {
@@ -470,41 +489,4 @@ TEST_CASE("varopt sketch: estimate subset sum", "[var_opt_sketch]") {
470
489
  REQUIRE(summary.estimate < total_weight); // exact mode, so know it must be strictly less
471
490
  }
472
491
 
473
- TEST_CASE("varopt sketch: deserialize exact from java", "[var_opt_sketch]") {
474
- std::ifstream is;
475
- is.exceptions(std::ios::failbit | std::ios::badbit);
476
- is.open(testBinaryInputPath + "varopt_sketch_string_exact.sk", std::ios::binary);
477
- var_opt_sketch<std::string> sketch = var_opt_sketch<std::string>::deserialize(is);
478
- REQUIRE_FALSE(sketch.is_empty());
479
- REQUIRE(sketch.get_k() == 1024);
480
- REQUIRE(sketch.get_n() == 200);
481
- REQUIRE(sketch.get_num_samples() == 200);
482
- subset_summary ss = sketch.estimate_subset_sum([](std::string){ return true; });
483
-
484
- double tgt_wt = 0.0;
485
- for (int i = 1; i <= 200; ++i) { tgt_wt += 1000.0 / i; }
486
- REQUIRE(ss.total_sketch_weight == Approx(tgt_wt).margin(EPS));
487
- }
488
-
489
-
490
- TEST_CASE("varopt sketch: deserialize sampling from java", "[var_opt_sketch]") {
491
- std::ifstream is;
492
- is.exceptions(std::ios::failbit | std::ios::badbit);
493
- is.open(testBinaryInputPath + "varopt_sketch_long_sampling.sk", std::ios::binary);
494
- var_opt_sketch<int64_t> sketch = var_opt_sketch<int64_t>::deserialize(is);
495
- REQUIRE_FALSE(sketch.is_empty());
496
- REQUIRE(sketch.get_k() == 1024);
497
- REQUIRE(sketch.get_n() == 2003);
498
- REQUIRE(sketch.get_num_samples() == sketch.get_k());
499
- subset_summary ss = sketch.estimate_subset_sum([](int64_t){ return true; });
500
- REQUIRE(ss.estimate == Approx(332000.0).margin(EPS));
501
- REQUIRE(ss.total_sketch_weight == Approx(332000.0).margin(EPS));
502
-
503
- ss = sketch.estimate_subset_sum([](int64_t x){ return x < 0; });
504
- REQUIRE(ss.estimate == 330000.0); // heavy item, weight is exact
505
-
506
- ss = sketch.estimate_subset_sum([](int64_t x){ return x >= 0; });
507
- REQUIRE(ss.estimate == Approx(2000.0).margin(EPS));
508
- }
509
-
510
492
  }
@@ -0,0 +1,50 @@
1
+ /*
2
+ * Licensed to the Apache Software Foundation (ASF) under one
3
+ * or more contributor license agreements. See the NOTICE file
4
+ * distributed with this work for additional information
5
+ * regarding copyright ownership. The ASF licenses this file
6
+ * to you under the Apache License, Version 2.0 (the
7
+ * "License"); you may not use this file except in compliance
8
+ * with the License. You may obtain a copy of the License at
9
+ *
10
+ * http://www.apache.org/licenses/LICENSE-2.0
11
+ *
12
+ * Unless required by applicable law or agreed to in writing,
13
+ * software distributed under the License is distributed on an
14
+ * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
15
+ * KIND, either express or implied. See the License for the
16
+ * specific language governing permissions and limitations
17
+ * under the License.
18
+ */
19
+
20
+ #include <catch2/catch.hpp>
21
+ #include <fstream>
22
+ #include <var_opt_union.hpp>
23
+
24
+ namespace datasketches {
25
+
26
+ // assume the binary sketches for this test have been generated by datasketches-java code
27
+ // in the subdirectory called "java" in the root directory of this project
28
+ static std::string testBinaryInputPath = std::string(TEST_BINARY_INPUT_PATH) + "../../java/";
29
+
30
+ TEST_CASE("var opt union double", "[serde_compat]") {
31
+ const double EPS = 1e-13;
32
+
33
+ std::ifstream is;
34
+ is.exceptions(std::ios::failbit | std::ios::badbit);
35
+ is.open(testBinaryInputPath + "varopt_union_double_sampling_java.sk", std::ios::binary);
36
+ auto u = var_opt_union<double>::deserialize(is);
37
+
38
+ // must reduce k in the process
39
+ const auto result = u.get_result();
40
+ REQUIRE_FALSE(result.is_empty());
41
+ REQUIRE(result.get_n() == 97);
42
+
43
+ const double expected_wt = 96.0; // light items -- ignoring the heavy one
44
+ const subset_summary ss = result.estimate_subset_sum([](double x){return x >= 0;});
45
+ REQUIRE(ss.estimate == Approx(expected_wt).margin(EPS));
46
+ REQUIRE(ss.total_sketch_weight == Approx(expected_wt + 1024.0).margin(EPS));
47
+ REQUIRE(result.get_k() < 128);
48
+ }
49
+
50
+ } /* namespace datasketches */
@@ -0,0 +1,56 @@
1
+ /*
2
+ * Licensed to the Apache Software Foundation (ASF) under one
3
+ * or more contributor license agreements. See the NOTICE file
4
+ * distributed with this work for additional information
5
+ * regarding copyright ownership. The ASF licenses this file
6
+ * to you under the Apache License, Version 2.0 (the
7
+ * "License"); you may not use this file except in compliance
8
+ * with the License. You may obtain a copy of the License at
9
+ *
10
+ * http://www.apache.org/licenses/LICENSE-2.0
11
+ *
12
+ * Unless required by applicable law or agreed to in writing,
13
+ * software distributed under the License is distributed on an
14
+ * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
15
+ * KIND, either express or implied. See the License for the
16
+ * specific language governing permissions and limitations
17
+ * under the License.
18
+ */
19
+
20
+ #include <catch2/catch.hpp>
21
+ #include <fstream>
22
+ #include <var_opt_union.hpp>
23
+
24
+ namespace datasketches {
25
+
26
+ TEST_CASE("var opt union double sampling", "[serialize_for_java]") {
27
+ const unsigned k_small = 16;
28
+ const unsigned k_max = 128;
29
+ const unsigned n1 = 32;
30
+ const unsigned n2 = 64;
31
+
32
+ // small k sketch, but sampling
33
+ var_opt_sketch<double> sketch1(k_small);
34
+ for (unsigned i = 0; i < n1; ++i) sketch1.update(i);
35
+ // negative heavy item to allow a simple predicate to filter
36
+ sketch1.update(-1, n1 * n1);
37
+
38
+ // another one, but different n to get a different per-item weight
39
+ var_opt_sketch<double> sketch2(k_small);
40
+ for (unsigned i = 0; i < n2; ++i) sketch2.update(i);
41
+
42
+ var_opt_union<double> u(k_max);
43
+ u.update(sketch1);
44
+ u.update(sketch2);
45
+
46
+ // must reduce k in the process
47
+ auto result = u.get_result();
48
+ REQUIRE(result.get_k() < k_max);
49
+ REQUIRE(result.get_k() >= k_small);
50
+ REQUIRE(result.get_n() == 97);
51
+
52
+ std::ofstream os("varopt_union_double_sampling_cpp.sk", std::ios::binary);
53
+ u.serialize(os);
54
+ }
55
+
56
+ } /* namespace datasketches */
@@ -57,7 +57,6 @@ static void check_if_equal(var_opt_sketch<T, A>& sk1, var_opt_sketch<T, A>& sk2,
57
57
 
58
58
  auto it1 = sk1.begin();
59
59
  auto it2 = sk2.begin();
60
- size_t i = 0;
61
60
 
62
61
  while ((it1 != sk1.end()) && (it2 != sk2.end())) {
63
62
  const std::pair<const T&, const double> p1 = *it1;
@@ -66,7 +65,6 @@ static void check_if_equal(var_opt_sketch<T, A>& sk1, var_opt_sketch<T, A>& sk2,
66
65
  REQUIRE(p1.first == p2.first); // data values
67
66
  }
68
67
  REQUIRE(p1.second == p2.second); // weight values
69
- ++i;
70
68
  ++it1;
71
69
  ++it2;
72
70
  }
@@ -307,22 +305,4 @@ TEST_CASE("varopt union: serialize sampling", "[var_opt_union]") {
307
305
  compare_serialization_deserialization(u);
308
306
  }
309
307
 
310
- TEST_CASE("varopt union: deserialize from java", "[var_opt_union]") {
311
- std::ifstream is;
312
- is.exceptions(std::ios::failbit | std::ios::badbit);
313
- is.open(testBinaryInputPath + "varopt_union_double_sampling.sk", std::ios::binary);
314
- var_opt_union<double> u = var_opt_union<double>::deserialize(is);
315
-
316
- // must reduce k in the process, like in small_sampling_sketch()
317
- var_opt_sketch<double> result = u.get_result();
318
- REQUIRE_FALSE(result.is_empty());
319
- REQUIRE(result.get_n() == 97);
320
-
321
- double expected_wt = 96.0;// light items -- ignoring the heavy one
322
- subset_summary ss = result.estimate_subset_sum([](double x){return x >= 0;});
323
- REQUIRE(ss.estimate == Approx(expected_wt).margin(EPS));
324
- REQUIRE(ss.total_sketch_weight == Approx(expected_wt + 1024.0).margin(EPS));
325
- REQUIRE(result.get_k() < 128);
326
- }
327
-
328
308
  }
@@ -62,4 +62,5 @@ install(FILES
62
62
  include/bounds_on_ratios_in_theta_sketched_sets.hpp
63
63
  include/compact_theta_sketch_parser.hpp
64
64
  include/compact_theta_sketch_parser_impl.hpp
65
+ include/bit_packing.hpp
65
66
  DESTINATION "${CMAKE_INSTALL_INCLUDEDIR}/DataSketches")