datasketches 0.1.1 → 0.2.2

Sign up to get free protection for your applications and to get access to all the features.
Files changed (197) hide show
  1. checksums.yaml +4 -4
  2. data/CHANGELOG.md +17 -0
  3. data/README.md +80 -4
  4. data/ext/datasketches/cpc_wrapper.cpp +34 -17
  5. data/ext/datasketches/ext.cpp +1 -1
  6. data/ext/datasketches/ext.h +4 -0
  7. data/ext/datasketches/extconf.rb +1 -1
  8. data/ext/datasketches/fi_wrapper.cpp +6 -8
  9. data/ext/datasketches/hll_wrapper.cpp +30 -22
  10. data/ext/datasketches/kll_wrapper.cpp +54 -45
  11. data/ext/datasketches/theta_wrapper.cpp +81 -5
  12. data/ext/datasketches/vo_wrapper.cpp +13 -13
  13. data/lib/datasketches/version.rb +1 -1
  14. data/vendor/datasketches-cpp/CMakeLists.txt +8 -0
  15. data/vendor/datasketches-cpp/README.md +4 -4
  16. data/vendor/datasketches-cpp/common/include/MurmurHash3.h +18 -7
  17. data/vendor/datasketches-cpp/common/include/binomial_bounds.hpp +8 -8
  18. data/vendor/datasketches-cpp/common/include/bounds_binomial_proportions.hpp +12 -15
  19. data/vendor/datasketches-cpp/common/include/common_defs.hpp +24 -0
  20. data/vendor/datasketches-cpp/common/include/conditional_forward.hpp +20 -8
  21. data/vendor/datasketches-cpp/common/include/count_zeros.hpp +2 -2
  22. data/vendor/datasketches-cpp/common/include/memory_operations.hpp +12 -0
  23. data/vendor/datasketches-cpp/common/include/serde.hpp +7 -7
  24. data/vendor/datasketches-cpp/common/test/CMakeLists.txt +24 -0
  25. data/vendor/datasketches-cpp/common/test/integration_test.cpp +77 -0
  26. data/vendor/datasketches-cpp/common/test/test_allocator.hpp +9 -1
  27. data/vendor/datasketches-cpp/cpc/include/cpc_common.hpp +3 -0
  28. data/vendor/datasketches-cpp/cpc/include/cpc_compressor.hpp +20 -20
  29. data/vendor/datasketches-cpp/cpc/include/cpc_compressor_impl.hpp +116 -105
  30. data/vendor/datasketches-cpp/cpc/include/cpc_sketch.hpp +22 -6
  31. data/vendor/datasketches-cpp/cpc/include/cpc_sketch_impl.hpp +140 -101
  32. data/vendor/datasketches-cpp/cpc/include/cpc_union.hpp +2 -2
  33. data/vendor/datasketches-cpp/cpc/include/cpc_union_impl.hpp +20 -20
  34. data/vendor/datasketches-cpp/cpc/include/cpc_util.hpp +10 -16
  35. data/vendor/datasketches-cpp/cpc/include/icon_estimator.hpp +6 -6
  36. data/vendor/datasketches-cpp/cpc/include/u32_table.hpp +10 -10
  37. data/vendor/datasketches-cpp/cpc/include/u32_table_impl.hpp +21 -21
  38. data/vendor/datasketches-cpp/cpc/test/CMakeLists.txt +1 -0
  39. data/vendor/datasketches-cpp/cpc/test/compression_test.cpp +10 -10
  40. data/vendor/datasketches-cpp/cpc/test/cpc_sketch_allocation_test.cpp +237 -0
  41. data/vendor/datasketches-cpp/cpc/test/cpc_sketch_test.cpp +25 -0
  42. data/vendor/datasketches-cpp/cpc/test/cpc_union_test.cpp +1 -1
  43. data/vendor/datasketches-cpp/fi/include/frequent_items_sketch.hpp +15 -10
  44. data/vendor/datasketches-cpp/fi/include/frequent_items_sketch_impl.hpp +102 -105
  45. data/vendor/datasketches-cpp/fi/include/reverse_purge_hash_map.hpp +19 -13
  46. data/vendor/datasketches-cpp/fi/include/reverse_purge_hash_map_impl.hpp +141 -125
  47. data/vendor/datasketches-cpp/fi/test/frequent_items_sketch_custom_type_test.cpp +15 -12
  48. data/vendor/datasketches-cpp/fi/test/reverse_purge_hash_map_test.cpp +5 -5
  49. data/vendor/datasketches-cpp/hll/CMakeLists.txt +3 -0
  50. data/vendor/datasketches-cpp/hll/include/AuxHashMap-internal.hpp +81 -109
  51. data/vendor/datasketches-cpp/hll/include/AuxHashMap.hpp +25 -24
  52. data/vendor/datasketches-cpp/hll/include/CompositeInterpolationXTable-internal.hpp +15 -15
  53. data/vendor/datasketches-cpp/hll/include/CompositeInterpolationXTable.hpp +5 -5
  54. data/vendor/datasketches-cpp/hll/include/CouponHashSet-internal.hpp +89 -105
  55. data/vendor/datasketches-cpp/hll/include/CouponHashSet.hpp +13 -13
  56. data/vendor/datasketches-cpp/hll/include/CouponList-internal.hpp +130 -165
  57. data/vendor/datasketches-cpp/hll/include/CouponList.hpp +21 -22
  58. data/vendor/datasketches-cpp/hll/include/CubicInterpolation-internal.hpp +2 -4
  59. data/vendor/datasketches-cpp/hll/include/CubicInterpolation.hpp +2 -2
  60. data/vendor/datasketches-cpp/hll/include/HarmonicNumbers-internal.hpp +1 -1
  61. data/vendor/datasketches-cpp/hll/include/HarmonicNumbers.hpp +2 -2
  62. data/vendor/datasketches-cpp/hll/include/Hll4Array-internal.hpp +88 -83
  63. data/vendor/datasketches-cpp/hll/include/Hll4Array.hpp +9 -9
  64. data/vendor/datasketches-cpp/hll/include/Hll6Array-internal.hpp +34 -45
  65. data/vendor/datasketches-cpp/hll/include/Hll6Array.hpp +7 -8
  66. data/vendor/datasketches-cpp/hll/include/Hll8Array-internal.hpp +41 -52
  67. data/vendor/datasketches-cpp/hll/include/Hll8Array.hpp +7 -8
  68. data/vendor/datasketches-cpp/hll/include/HllArray-internal.hpp +220 -251
  69. data/vendor/datasketches-cpp/hll/include/HllArray.hpp +42 -42
  70. data/vendor/datasketches-cpp/hll/include/HllSketch-internal.hpp +36 -38
  71. data/vendor/datasketches-cpp/hll/include/HllSketchImpl-internal.hpp +22 -22
  72. data/vendor/datasketches-cpp/hll/include/HllSketchImpl.hpp +15 -14
  73. data/vendor/datasketches-cpp/hll/include/HllSketchImplFactory.hpp +47 -44
  74. data/vendor/datasketches-cpp/hll/include/HllUnion-internal.hpp +62 -87
  75. data/vendor/datasketches-cpp/hll/include/HllUtil.hpp +121 -128
  76. data/vendor/datasketches-cpp/hll/include/RelativeErrorTables.hpp +1 -1
  77. data/vendor/datasketches-cpp/hll/include/coupon_iterator-internal.hpp +9 -9
  78. data/vendor/datasketches-cpp/hll/include/coupon_iterator.hpp +5 -5
  79. data/vendor/datasketches-cpp/hll/include/hll.hpp +25 -53
  80. data/vendor/datasketches-cpp/hll/test/AuxHashMapTest.cpp +8 -8
  81. data/vendor/datasketches-cpp/hll/test/CouponHashSetTest.cpp +36 -36
  82. data/vendor/datasketches-cpp/hll/test/CouponListTest.cpp +28 -28
  83. data/vendor/datasketches-cpp/hll/test/CrossCountingTest.cpp +2 -2
  84. data/vendor/datasketches-cpp/hll/test/HllArrayTest.cpp +37 -37
  85. data/vendor/datasketches-cpp/hll/test/HllSketchTest.cpp +57 -61
  86. data/vendor/datasketches-cpp/hll/test/HllUnionTest.cpp +10 -14
  87. data/vendor/datasketches-cpp/hll/test/IsomorphicTest.cpp +3 -3
  88. data/vendor/datasketches-cpp/hll/test/ToFromByteArrayTest.cpp +4 -4
  89. data/vendor/datasketches-cpp/kll/include/kll_helper.hpp +5 -4
  90. data/vendor/datasketches-cpp/kll/include/kll_helper_impl.hpp +6 -6
  91. data/vendor/datasketches-cpp/kll/include/kll_quantile_calculator.hpp +14 -6
  92. data/vendor/datasketches-cpp/kll/include/kll_quantile_calculator_impl.hpp +40 -25
  93. data/vendor/datasketches-cpp/kll/include/kll_sketch.hpp +44 -5
  94. data/vendor/datasketches-cpp/kll/include/kll_sketch_impl.hpp +164 -136
  95. data/vendor/datasketches-cpp/kll/include/kolmogorov_smirnov.hpp +67 -0
  96. data/vendor/datasketches-cpp/kll/include/kolmogorov_smirnov_impl.hpp +78 -0
  97. data/vendor/datasketches-cpp/kll/test/CMakeLists.txt +1 -0
  98. data/vendor/datasketches-cpp/kll/test/kll_sketch_custom_type_test.cpp +11 -10
  99. data/vendor/datasketches-cpp/kll/test/kll_sketch_test.cpp +113 -87
  100. data/vendor/datasketches-cpp/kll/test/kolmogorov_smirnov_test.cpp +111 -0
  101. data/vendor/datasketches-cpp/python/CMakeLists.txt +2 -0
  102. data/vendor/datasketches-cpp/python/README.md +6 -3
  103. data/vendor/datasketches-cpp/python/src/datasketches.cpp +2 -0
  104. data/vendor/datasketches-cpp/python/src/hll_wrapper.cpp +4 -6
  105. data/vendor/datasketches-cpp/python/src/kll_wrapper.cpp +3 -1
  106. data/vendor/datasketches-cpp/python/src/req_wrapper.cpp +246 -0
  107. data/vendor/datasketches-cpp/python/src/theta_wrapper.cpp +36 -26
  108. data/vendor/datasketches-cpp/python/src/vo_wrapper.cpp +2 -2
  109. data/vendor/datasketches-cpp/python/tests/hll_test.py +1 -2
  110. data/vendor/datasketches-cpp/python/tests/kll_test.py +3 -3
  111. data/vendor/datasketches-cpp/python/tests/req_test.py +126 -0
  112. data/vendor/datasketches-cpp/python/tests/theta_test.py +28 -3
  113. data/vendor/datasketches-cpp/python/tests/vo_test.py +3 -3
  114. data/vendor/datasketches-cpp/req/CMakeLists.txt +60 -0
  115. data/vendor/datasketches-cpp/{tuple/include/theta_a_not_b_experimental_impl.hpp → req/include/req_common.hpp} +18 -8
  116. data/vendor/datasketches-cpp/req/include/req_compactor.hpp +137 -0
  117. data/vendor/datasketches-cpp/req/include/req_compactor_impl.hpp +488 -0
  118. data/vendor/datasketches-cpp/req/include/req_quantile_calculator.hpp +69 -0
  119. data/vendor/datasketches-cpp/req/include/req_quantile_calculator_impl.hpp +60 -0
  120. data/vendor/datasketches-cpp/req/include/req_sketch.hpp +395 -0
  121. data/vendor/datasketches-cpp/req/include/req_sketch_impl.hpp +810 -0
  122. data/vendor/datasketches-cpp/req/test/CMakeLists.txt +43 -0
  123. data/vendor/datasketches-cpp/req/test/req_float_empty_from_java.sk +0 -0
  124. data/vendor/datasketches-cpp/req/test/req_float_estimation_from_java.sk +0 -0
  125. data/vendor/datasketches-cpp/req/test/req_float_exact_from_java.sk +0 -0
  126. data/vendor/datasketches-cpp/req/test/req_float_raw_items_from_java.sk +0 -0
  127. data/vendor/datasketches-cpp/req/test/req_float_single_item_from_java.sk +0 -0
  128. data/vendor/datasketches-cpp/req/test/req_sketch_custom_type_test.cpp +128 -0
  129. data/vendor/datasketches-cpp/req/test/req_sketch_test.cpp +494 -0
  130. data/vendor/datasketches-cpp/sampling/include/var_opt_sketch.hpp +10 -9
  131. data/vendor/datasketches-cpp/sampling/include/var_opt_sketch_impl.hpp +129 -126
  132. data/vendor/datasketches-cpp/sampling/include/var_opt_union.hpp +5 -5
  133. data/vendor/datasketches-cpp/sampling/include/var_opt_union_impl.hpp +41 -49
  134. data/vendor/datasketches-cpp/sampling/test/CMakeLists.txt +1 -0
  135. data/vendor/datasketches-cpp/sampling/test/var_opt_allocation_test.cpp +96 -0
  136. data/vendor/datasketches-cpp/sampling/test/var_opt_sketch_test.cpp +6 -6
  137. data/vendor/datasketches-cpp/sampling/test/var_opt_union_test.cpp +13 -44
  138. data/vendor/datasketches-cpp/setup.py +5 -3
  139. data/vendor/datasketches-cpp/theta/CMakeLists.txt +30 -3
  140. data/vendor/datasketches-cpp/{tuple → theta}/include/bounds_on_ratios_in_sampled_sets.hpp +3 -2
  141. data/vendor/datasketches-cpp/{tuple → theta}/include/bounds_on_ratios_in_theta_sketched_sets.hpp +1 -1
  142. data/vendor/datasketches-cpp/theta/include/compact_theta_sketch_parser.hpp +67 -0
  143. data/vendor/datasketches-cpp/theta/include/compact_theta_sketch_parser_impl.hpp +70 -0
  144. data/vendor/datasketches-cpp/theta/include/theta_a_not_b.hpp +12 -29
  145. data/vendor/datasketches-cpp/theta/include/theta_a_not_b_impl.hpp +5 -46
  146. data/vendor/datasketches-cpp/{tuple → theta}/include/theta_comparators.hpp +0 -0
  147. data/vendor/datasketches-cpp/{tuple → theta}/include/theta_constants.hpp +2 -0
  148. data/vendor/datasketches-cpp/{tuple → theta}/include/theta_helpers.hpp +0 -0
  149. data/vendor/datasketches-cpp/theta/include/theta_intersection.hpp +26 -28
  150. data/vendor/datasketches-cpp/{tuple → theta}/include/theta_intersection_base.hpp +0 -0
  151. data/vendor/datasketches-cpp/{tuple → theta}/include/theta_intersection_base_impl.hpp +0 -0
  152. data/vendor/datasketches-cpp/theta/include/theta_intersection_impl.hpp +8 -90
  153. data/vendor/datasketches-cpp/{tuple/test/theta_union_experimental_test.cpp → theta/include/theta_jaccard_similarity.hpp} +11 -18
  154. data/vendor/datasketches-cpp/{tuple/include/jaccard_similarity.hpp → theta/include/theta_jaccard_similarity_base.hpp} +24 -36
  155. data/vendor/datasketches-cpp/{tuple → theta}/include/theta_set_difference_base.hpp +0 -0
  156. data/vendor/datasketches-cpp/{tuple → theta}/include/theta_set_difference_base_impl.hpp +5 -0
  157. data/vendor/datasketches-cpp/theta/include/theta_sketch.hpp +163 -256
  158. data/vendor/datasketches-cpp/theta/include/theta_sketch_impl.hpp +250 -651
  159. data/vendor/datasketches-cpp/theta/include/theta_union.hpp +27 -60
  160. data/vendor/datasketches-cpp/{tuple → theta}/include/theta_union_base.hpp +1 -1
  161. data/vendor/datasketches-cpp/{tuple → theta}/include/theta_union_base_impl.hpp +6 -1
  162. data/vendor/datasketches-cpp/theta/include/theta_union_impl.hpp +13 -69
  163. data/vendor/datasketches-cpp/{tuple → theta}/include/theta_update_sketch_base.hpp +5 -19
  164. data/vendor/datasketches-cpp/{tuple → theta}/include/theta_update_sketch_base_impl.hpp +39 -29
  165. data/vendor/datasketches-cpp/theta/test/CMakeLists.txt +1 -0
  166. data/vendor/datasketches-cpp/theta/test/theta_a_not_b_test.cpp +23 -1
  167. data/vendor/datasketches-cpp/theta/test/theta_intersection_test.cpp +21 -1
  168. data/vendor/datasketches-cpp/{tuple → theta}/test/theta_jaccard_similarity_test.cpp +60 -5
  169. data/vendor/datasketches-cpp/theta/test/theta_sketch_test.cpp +74 -235
  170. data/vendor/datasketches-cpp/theta/test/theta_union_test.cpp +22 -2
  171. data/vendor/datasketches-cpp/tuple/CMakeLists.txt +3 -35
  172. data/vendor/datasketches-cpp/tuple/include/array_of_doubles_sketch_impl.hpp +47 -60
  173. data/vendor/datasketches-cpp/tuple/include/tuple_jaccard_similarity.hpp +38 -0
  174. data/vendor/datasketches-cpp/tuple/include/tuple_sketch.hpp +28 -13
  175. data/vendor/datasketches-cpp/tuple/include/tuple_sketch_impl.hpp +57 -70
  176. data/vendor/datasketches-cpp/tuple/test/CMakeLists.txt +1 -6
  177. data/vendor/datasketches-cpp/tuple/test/array_of_doubles_sketch_test.cpp +1 -1
  178. data/vendor/datasketches-cpp/tuple/test/tuple_a_not_b_test.cpp +18 -21
  179. data/vendor/datasketches-cpp/tuple/test/tuple_intersection_test.cpp +13 -16
  180. data/vendor/datasketches-cpp/tuple/test/tuple_jaccard_similarity_test.cpp +7 -6
  181. data/vendor/datasketches-cpp/tuple/test/tuple_sketch_allocation_test.cpp +3 -3
  182. data/vendor/datasketches-cpp/tuple/test/tuple_sketch_test.cpp +20 -20
  183. data/vendor/datasketches-cpp/tuple/test/tuple_union_test.cpp +13 -16
  184. metadata +50 -36
  185. data/vendor/datasketches-cpp/tuple/include/theta_a_not_b_experimental.hpp +0 -53
  186. data/vendor/datasketches-cpp/tuple/include/theta_intersection_experimental.hpp +0 -78
  187. data/vendor/datasketches-cpp/tuple/include/theta_intersection_experimental_impl.hpp +0 -43
  188. data/vendor/datasketches-cpp/tuple/include/theta_sketch_experimental.hpp +0 -393
  189. data/vendor/datasketches-cpp/tuple/include/theta_sketch_experimental_impl.hpp +0 -481
  190. data/vendor/datasketches-cpp/tuple/include/theta_union_experimental.hpp +0 -88
  191. data/vendor/datasketches-cpp/tuple/include/theta_union_experimental_impl.hpp +0 -47
  192. data/vendor/datasketches-cpp/tuple/test/theta_a_not_b_experimental_test.cpp +0 -250
  193. data/vendor/datasketches-cpp/tuple/test/theta_compact_empty_from_java.sk +0 -0
  194. data/vendor/datasketches-cpp/tuple/test/theta_compact_estimation_from_java.sk +0 -0
  195. data/vendor/datasketches-cpp/tuple/test/theta_compact_single_item_from_java.sk +0 -0
  196. data/vendor/datasketches-cpp/tuple/test/theta_intersection_experimental_test.cpp +0 -224
  197. data/vendor/datasketches-cpp/tuple/test/theta_sketch_experimental_test.cpp +0 -247
checksums.yaml CHANGED
@@ -1,7 +1,7 @@
1
1
  ---
2
2
  SHA256:
3
- metadata.gz: bf7f5b6898a9f014d14fc5fcfa3a6a5da1dd5fc70d1b07bbbdf07a446126d9cd
4
- data.tar.gz: 15d5fbb2ee6e0c1a6e7f25a145329eb73da9d7933f5f1475596b9e9c998b31f7
3
+ metadata.gz: 5c578044053c564421893cc4433f7fe557f23ba9d8a1995fc2a2c5f07742721a
4
+ data.tar.gz: f4122bd75e19fede015b01a5e5ad8e6130f75babe9c9160cc56f378480a16cee
5
5
  SHA512:
6
- metadata.gz: e3b412132bdaf7e45f85e1da48690ecccd3c510cc17fd5259f0244cb580f712dc58290fca50a547a9e0952cf7c0548124ed33be9224ba3b83d6dd461d70c2aab
7
- data.tar.gz: 5f6d9932824961e11d6f8bd390dbf20e487b6ca8aef6b6440e5f6a11296227382be5261ebaf1da416d5f2354c7c99a97e34871b0964facc724ba33f2b88e96f6
6
+ metadata.gz: 2d7c4d7306f28356557a816a78033b909561ccd8f843281a2b756e88cbdcb9936da7995ff80871a19e229675ead812aca00d6c639d63a6532998c3c1b35aa953
7
+ data.tar.gz: fdf0fe1d14e04bfddef9df1ae7958f6571a7f689865aa02e81713d4b250afeeeb8c90a168ce855728d9c831aff6d3ea71df91c71b7269a760c19488c42c92658
data/CHANGELOG.md CHANGED
@@ -1,3 +1,20 @@
1
+ ## 0.2.2 (2021-07-17)
2
+
3
+ - Updated DataSketches to 3.1.0
4
+
5
+ ## 0.2.1 (2021-05-23)
6
+
7
+ - Improved performance
8
+
9
+ ## 0.2.0 (2021-05-17)
10
+
11
+ - Updated DataSketches to 3.0.0
12
+ - Updated to Rice 4
13
+
14
+ ## 0.1.2 (2021-01-21)
15
+
16
+ - Added `CpcUnion` and `UpdateThetaSketch`
17
+
1
18
  ## 0.1.1 (2021-01-20)
2
19
 
3
20
  - Added more sketches
data/README.md CHANGED
@@ -18,6 +18,7 @@ Distinct counting
18
18
 
19
19
  - [CPC sketch](#cpc-sketch)
20
20
  - [HyperLogLog sketch](#hyperloglog-sketch)
21
+ - [Theta sketch](#theta-sketch)
21
22
 
22
23
  Most frequent
23
24
 
@@ -36,7 +37,7 @@ Sampling
36
37
  Create a sketch
37
38
 
38
39
  ```ruby
39
- sketch = DataSketches::CpcSketch.new(14)
40
+ sketch = DataSketches::CpcSketch.new
40
41
  ```
41
42
 
42
43
  Add data
@@ -65,6 +66,15 @@ Load a sketch
65
66
  sketch = DataSketches::CpcSketch.deserialize(data)
66
67
  ```
67
68
 
69
+ Get the union
70
+
71
+ ```ruby
72
+ u = DataSketches::CpcUnion.new(14)
73
+ u.update(sketch1)
74
+ u.update(sketch2)
75
+ u.result
76
+ ```
77
+
68
78
  ## HyperLogLog Sketch
69
79
 
70
80
  Create a sketch
@@ -107,7 +117,66 @@ Get the union
107
117
  u = DataSketches::HllUnion.new(14)
108
118
  u.update(sketch1)
109
119
  u.update(sketch2)
110
- u.estimate
120
+ u.result
121
+ ```
122
+
123
+ ## Theta Sketch
124
+
125
+ Create a sketch
126
+
127
+ ```ruby
128
+ sketch = DataSketches::UpdateThetaSketch.new
129
+ ```
130
+
131
+ Add data
132
+
133
+ ```ruby
134
+ sketch.update(1)
135
+ sketch.update(2.0)
136
+ sketch.update("three")
137
+ ```
138
+
139
+ Estimate the count
140
+
141
+ ```ruby
142
+ sketch.estimate
143
+ ```
144
+
145
+ Save a sketch
146
+
147
+ ```ruby
148
+ data = sketch.serialize
149
+ ```
150
+
151
+ Load a sketch
152
+
153
+ ```ruby
154
+ sketch = DataSketches::UpdateThetaSketch.deserialize(data)
155
+ ```
156
+
157
+ Get the union
158
+
159
+ ```ruby
160
+ u = DataSketches::ThetaUnion.new
161
+ u.update(sketch1)
162
+ u.update(sketch2)
163
+ u.result
164
+ ```
165
+
166
+ Get the intersection
167
+
168
+ ```ruby
169
+ i = DataSketches::ThetaIntersection.new
170
+ i.update(sketch1)
171
+ i.update(sketch2)
172
+ i.result
173
+ ```
174
+
175
+ Compute A not B
176
+
177
+ ```ruby
178
+ d = DataSketches::ThetaANotB.new
179
+ d.compute(a, b)
111
180
  ```
112
181
 
113
182
  ## Frequent Item Sketch
@@ -149,9 +218,9 @@ sketch = DataSketches::FrequentStringsSketch.deserialize(data)
149
218
  Create a sketch
150
219
 
151
220
  ```ruby
152
- sketch = DataSketches::KllIntsSketch.new(200)
221
+ sketch = DataSketches::KllIntsSketch.new
153
222
  # or
154
- sketch = DataSketches::KllFloatsSketch.new(200)
223
+ sketch = DataSketches::KllFloatsSketch.new
155
224
  ```
156
225
 
157
226
  Add data
@@ -167,6 +236,7 @@ Get quantiles
167
236
  ```ruby
168
237
  sketch.quantile(0.5)
169
238
  sketch.quantiles([0.25, 0.5, 0.75])
239
+ sketch.quantiles(3)
170
240
  ```
171
241
 
172
242
  Get the minimum and maximum values from the stream
@@ -210,6 +280,12 @@ sketch.update(2.0)
210
280
  sketch.update("three")
211
281
  ```
212
282
 
283
+ Sample data
284
+
285
+ ```ruby
286
+ sketch.samples
287
+ ```
288
+
213
289
  ## Credits
214
290
 
215
291
  This library is modeled after the DataSketches [Python API](https://github.com/apache/datasketches-cpp/tree/master/python).
@@ -1,36 +1,44 @@
1
1
  #include <sstream>
2
2
 
3
3
  #include <cpc_sketch.hpp>
4
+ #include <cpc_union.hpp>
4
5
 
5
- #include <rice/Constructor.hpp>
6
- #include <rice/Module.hpp>
6
+ #include "ext.h"
7
+
8
+ using datasketches::cpc_sketch;
9
+ using datasketches::cpc_union;
10
+
11
+ using datasketches::CPC_DEFAULT_LG_K;
12
+ using datasketches::DEFAULT_SEED;
13
+
14
+ using Rice::Arg;
7
15
 
8
16
  void init_cpc(Rice::Module& m) {
9
- Rice::define_class_under<datasketches::cpc_sketch>(m, "CpcSketch")
10
- .define_constructor(Rice::Constructor<datasketches::cpc_sketch, uint8_t>())
11
- .define_method("lg_k", &datasketches::cpc_sketch::get_lg_k)
12
- .define_method("empty?", &datasketches::cpc_sketch::is_empty)
13
- .define_method("lower_bound", &datasketches::cpc_sketch::get_lower_bound)
14
- .define_method("upper_bound", &datasketches::cpc_sketch::get_upper_bound)
17
+ Rice::define_class_under<cpc_sketch>(m, "CpcSketch")
18
+ .define_constructor(Rice::Constructor<cpc_sketch, uint8_t, uint64_t>(), Rice::Arg("lg_k")=CPC_DEFAULT_LG_K, Rice::Arg("seed")=DEFAULT_SEED)
19
+ .define_method("lg_k", &cpc_sketch::get_lg_k)
20
+ .define_method("empty?", &cpc_sketch::is_empty)
21
+ .define_method("lower_bound", &cpc_sketch::get_lower_bound)
22
+ .define_method("upper_bound", &cpc_sketch::get_upper_bound)
15
23
  .define_method(
16
24
  "update",
17
- *[](datasketches::cpc_sketch& self, Rice::Object datum) {
25
+ [](cpc_sketch& self, Rice::Object datum) {
18
26
  if (FIXNUM_P(datum.value())) {
19
- return self.update(from_ruby<int64_t>(datum));
27
+ return self.update(Rice::detail::From_Ruby<int64_t>().convert(datum));
20
28
  } else if (datum.is_a(rb_cNumeric)) {
21
- return self.update(from_ruby<double>(datum));
29
+ return self.update(Rice::detail::From_Ruby<double>().convert(datum));
22
30
  } else {
23
31
  return self.update(datum.to_s().str());
24
32
  }
25
33
  })
26
34
  .define_method(
27
35
  "estimate",
28
- *[](datasketches::cpc_sketch& self) {
36
+ [](cpc_sketch& self) {
29
37
  return self.get_estimate();
30
38
  })
31
39
  .define_method(
32
40
  "serialize",
33
- *[](datasketches::cpc_sketch& self) {
41
+ [](cpc_sketch& self) {
34
42
  std::ostringstream oss;
35
43
  self.serialize(oss);
36
44
  return oss.str();
@@ -38,13 +46,22 @@ void init_cpc(Rice::Module& m) {
38
46
  // TODO change to summary?
39
47
  .define_method(
40
48
  "to_string",
41
- *[](datasketches::cpc_sketch& self) {
49
+ [](cpc_sketch& self) {
42
50
  return self.to_string();
43
51
  })
44
- .define_singleton_method(
52
+ .define_singleton_function(
45
53
  "deserialize",
46
- *[](std::string& is) {
54
+ [](const std::string& is) {
47
55
  std::istringstream iss(is);
48
- return datasketches::cpc_sketch::deserialize(iss);
56
+ return cpc_sketch::deserialize(iss);
57
+ });
58
+
59
+ Rice::define_class_under<cpc_union>(m, "CpcUnion")
60
+ .define_constructor(Rice::Constructor<cpc_union, uint8_t, uint64_t>(), Rice::Arg("lg_k"), Rice::Arg("seed")=DEFAULT_SEED)
61
+ .define_method("result", &cpc_union::get_result)
62
+ .define_method(
63
+ "update",
64
+ [](cpc_union& self, cpc_sketch& sketch) {
65
+ self.update(sketch);
49
66
  });
50
67
  }
@@ -1,4 +1,4 @@
1
- #include <rice/Module.hpp>
1
+ #include "ext.h"
2
2
 
3
3
  void init_cpc(Rice::Module& m);
4
4
  void init_fi(Rice::Module& m);
@@ -0,0 +1,4 @@
1
+ #pragma once
2
+
3
+ #include <rice/rice.hpp>
4
+ #include <rice/stl.hpp>
@@ -1,6 +1,6 @@
1
1
  require "mkmf-rice"
2
2
 
3
- $CXXFLAGS += " -std=c++11"
3
+ $CXXFLAGS += " -std=c++17 $(optflags)"
4
4
 
5
5
  ext = File.expand_path(".", __dir__)
6
6
  datasketches = File.expand_path("../../vendor/datasketches-cpp", __dir__)
@@ -2,8 +2,7 @@
2
2
 
3
3
  #include <frequent_items_sketch.hpp>
4
4
 
5
- #include <rice/Constructor.hpp>
6
- #include <rice/Module.hpp>
5
+ #include "ext.h"
7
6
 
8
7
  template<typename T>
9
8
  void bind_fi_sketch(Rice::Module& m, const char* name) {
@@ -18,12 +17,12 @@ void bind_fi_sketch(Rice::Module& m, const char* name) {
18
17
  .define_method("maximum_error", &datasketches::frequent_items_sketch<T>::get_maximum_error)
19
18
  .define_method(
20
19
  "update",
21
- *[](datasketches::frequent_items_sketch<T>& self, const T item) {
20
+ [](datasketches::frequent_items_sketch<T>& self, const T item) {
22
21
  self.update(item);
23
22
  })
24
23
  .define_method(
25
24
  "serialize",
26
- *[](datasketches::frequent_items_sketch<T>& self) {
25
+ [](datasketches::frequent_items_sketch<T>& self) {
27
26
  std::ostringstream oss;
28
27
  self.serialize(oss);
29
28
  return oss.str();
@@ -31,13 +30,12 @@ void bind_fi_sketch(Rice::Module& m, const char* name) {
31
30
  // TODO change to summary?
32
31
  .define_method(
33
32
  "to_string",
34
- *[](datasketches::frequent_items_sketch<T>& self) {
33
+ [](datasketches::frequent_items_sketch<T>& self) {
35
34
  return self.to_string();
36
35
  })
37
- .define_singleton_method(
36
+ .define_singleton_function(
38
37
  "deserialize",
39
- // TODO figure out segfault
40
- *[](std::string is) {
38
+ [](const std::string& is) {
41
39
  std::istringstream iss(is);
42
40
  return datasketches::frequent_items_sketch<T>::deserialize(iss);
43
41
  });
@@ -2,41 +2,44 @@
2
2
 
3
3
  #include <hll.hpp>
4
4
 
5
- #include <rice/Constructor.hpp>
6
- #include <rice/Module.hpp>
5
+ #include "ext.h"
6
+
7
+ using datasketches::hll_sketch;
8
+ using datasketches::hll_union;
7
9
 
8
10
  void init_hll(Rice::Module& m) {
9
- Rice::define_class_under<datasketches::hll_sketch>(m, "HllSketch")
10
- .define_constructor(Rice::Constructor<datasketches::hll_sketch, int>())
11
- .define_method("lg_config_k", &datasketches::hll_sketch::get_lg_config_k)
12
- .define_method("compact?", &datasketches::hll_sketch::is_compact)
13
- .define_method("empty?", &datasketches::hll_sketch::is_empty)
11
+ Rice::define_class_under<hll_sketch>(m, "HllSketch")
12
+ .define_constructor(Rice::Constructor<hll_sketch, int>())
13
+ .define_method("lg_config_k", &hll_sketch::get_lg_config_k)
14
+ .define_method("compact?", &hll_sketch::is_compact)
15
+ .define_method("empty?", &hll_sketch::is_empty)
16
+ .define_method("composite_estimate", &hll_sketch::get_composite_estimate)
14
17
  .define_method(
15
18
  "update",
16
- *[](datasketches::hll_sketch& self, Rice::Object datum) {
19
+ [](hll_sketch& self, Rice::Object datum) {
17
20
  if (FIXNUM_P(datum.value())) {
18
- return self.update(from_ruby<int64_t>(datum));
21
+ return self.update(Rice::detail::From_Ruby<int64_t>().convert(datum));
19
22
  } else if (datum.is_a(rb_cNumeric)) {
20
- return self.update(from_ruby<double>(datum));
23
+ return self.update(Rice::detail::From_Ruby<double>().convert(datum));
21
24
  } else {
22
25
  return self.update(datum.to_s().str());
23
26
  }
24
27
  })
25
28
  .define_method(
26
29
  "estimate",
27
- *[](datasketches::hll_sketch& self) {
30
+ [](hll_sketch& self) {
28
31
  return self.get_estimate();
29
32
  })
30
33
  .define_method(
31
34
  "serialize_compact",
32
- *[](datasketches::hll_sketch& self) {
35
+ [](hll_sketch& self) {
33
36
  std::ostringstream oss;
34
37
  self.serialize_compact(oss);
35
38
  return oss.str();
36
39
  })
37
40
  .define_method(
38
41
  "serialize_updatable",
39
- *[](datasketches::hll_sketch& self) {
42
+ [](hll_sketch& self) {
40
43
  std::ostringstream oss;
41
44
  self.serialize_updatable(oss);
42
45
  return oss.str();
@@ -44,26 +47,31 @@ void init_hll(Rice::Module& m) {
44
47
  // TODO change to summary?
45
48
  .define_method(
46
49
  "to_string",
47
- *[](datasketches::hll_sketch& self) {
50
+ [](hll_sketch& self) {
48
51
  return self.to_string();
49
52
  })
50
- .define_singleton_method(
53
+ .define_singleton_function(
51
54
  "deserialize",
52
- *[](std::string& is) {
55
+ [](const std::string& is) {
53
56
  std::istringstream iss(is);
54
- return datasketches::hll_sketch::deserialize(iss);
57
+ return hll_sketch::deserialize(iss);
55
58
  });
56
59
 
57
- Rice::define_class_under<datasketches::hll_union>(m, "HllUnion")
58
- .define_constructor(Rice::Constructor<datasketches::hll_union, int>())
60
+ Rice::define_class_under<hll_union>(m, "HllUnion")
61
+ .define_constructor(Rice::Constructor<hll_union, int>())
59
62
  .define_method(
60
63
  "update",
61
- *[](datasketches::hll_union& self, datasketches::hll_sketch& datum) {
62
- self.update(datum);
64
+ [](hll_union& self, hll_sketch& sketch) {
65
+ self.update(sketch);
63
66
  })
64
67
  .define_method(
65
68
  "estimate",
66
- *[](datasketches::hll_union& self) {
69
+ [](hll_union& self) {
67
70
  return self.get_estimate();
71
+ })
72
+ .define_method(
73
+ "result",
74
+ [](hll_union& self) {
75
+ return self.get_result();
68
76
  });
69
77
  }
@@ -2,67 +2,76 @@
2
2
 
3
3
  #include <kll_sketch.hpp>
4
4
 
5
- #include <rice/Array.hpp>
6
- #include <rice/Constructor.hpp>
7
- #include <rice/Module.hpp>
5
+ #include "ext.h"
8
6
 
9
- template<>
10
- std::vector<double> from_ruby<std::vector<double>>(Rice::Object x)
11
- {
12
- auto a = Rice::Array(x);
13
- std::vector<double> vec(a.size());
14
- for (size_t i = 0; i < a.size(); i++) {
15
- vec[i] = from_ruby<double>(a[i]);
16
- }
17
- return vec;
18
- }
19
-
20
- template<>
21
- Rice::Object to_ruby<std::vector<int>>(std::vector<int> const & x)
22
- {
23
- auto a = Rice::Array();
24
- for (size_t i = 0; i < x.size(); i++) {
25
- a.push(x[i]);
26
- }
27
- return a;
28
- }
7
+ using datasketches::kll_sketch;
29
8
 
30
- template<>
31
- Rice::Object to_ruby<std::vector<float>>(std::vector<float> const & x)
9
+ namespace Rice::detail
32
10
  {
33
- auto a = Rice::Array();
34
- for (size_t i = 0; i < x.size(); i++) {
35
- a.push(x[i]);
36
- }
37
- return a;
11
+ template<typename T>
12
+ class To_Ruby<std::vector<T>>
13
+ {
14
+ public:
15
+ VALUE convert(std::vector<T> const & x)
16
+ {
17
+ auto a = rb_ary_new2(x.size());
18
+ for (const auto& v : x) {
19
+ detail::protect(rb_ary_push, a, To_Ruby<T>().convert(v));
20
+ }
21
+ return a;
22
+ }
23
+ };
38
24
  }
39
25
 
40
26
  template<typename T>
41
27
  void bind_kll_sketch(Rice::Module& m, const char* name) {
42
- Rice::define_class_under<datasketches::kll_sketch<T>>(m, name)
43
- .define_constructor(Rice::Constructor<datasketches::kll_sketch<T>, uint16_t>())
44
- .define_method("empty?", &datasketches::kll_sketch<T>::is_empty)
45
- .define_method("min_value", &datasketches::kll_sketch<T>::get_min_value)
46
- .define_method("max_value", &datasketches::kll_sketch<T>::get_max_value)
47
- .define_method("quantile", &datasketches::kll_sketch<T>::get_quantile)
28
+ Rice::define_class_under<kll_sketch<T>>(m, name)
29
+ .define_constructor(Rice::Constructor<kll_sketch<T>, uint16_t>(), Rice::Arg("k")=kll_sketch<T>::DEFAULT_K)
30
+ .define_method("empty?", &kll_sketch<T>::is_empty)
31
+ .define_method("n", &kll_sketch<T>::get_n)
32
+ .define_method("num_retained", &kll_sketch<T>::get_num_retained)
33
+ .define_method("estimation_mode?", &kll_sketch<T>::is_estimation_mode)
34
+ .define_method("min_value", &kll_sketch<T>::get_min_value)
35
+ .define_method("max_value", &kll_sketch<T>::get_max_value)
36
+ .define_method("quantile", &kll_sketch<T>::get_quantile)
48
37
  .define_method(
49
38
  "quantiles",
50
- *[](datasketches::kll_sketch<T>& self, std::vector<double> fractions) {
51
- return self.get_quantiles(&fractions[0], fractions.size());
39
+ [](kll_sketch<T>& self, Rice::Object obj) {
40
+ if (obj.is_a(rb_cArray)) {
41
+ auto fractions = Rice::detail::From_Ruby<std::vector<double>>().convert(obj);
42
+ return self.get_quantiles(&fractions[0], fractions.size());
43
+ } else {
44
+ return self.get_quantiles(Rice::detail::From_Ruby<size_t>().convert(obj));
45
+ }
46
+ })
47
+ .define_method(
48
+ "rank",
49
+ [](kll_sketch<T>& self, const T item) {
50
+ return self.get_rank(item);
51
+ })
52
+ .define_method(
53
+ "pmf",
54
+ [](kll_sketch<T>& self, std::vector<T> split_points) {
55
+ return self.get_PMF(&split_points[0], split_points.size());
56
+ })
57
+ .define_method(
58
+ "cdf",
59
+ [](kll_sketch<T>& self, std::vector<T> split_points) {
60
+ return self.get_CDF(&split_points[0], split_points.size());
52
61
  })
53
62
  .define_method(
54
63
  "merge",
55
- *[](datasketches::kll_sketch<T>& self, const datasketches::kll_sketch<T>& other) {
64
+ [](kll_sketch<T>& self, const kll_sketch<T>& other) {
56
65
  self.merge(other);
57
66
  })
58
67
  .define_method(
59
68
  "update",
60
- *[](datasketches::kll_sketch<T>& self, const T item) {
69
+ [](kll_sketch<T>& self, const T item) {
61
70
  self.update(item);
62
71
  })
63
72
  .define_method(
64
73
  "serialize",
65
- *[](datasketches::kll_sketch<T>& self) {
74
+ [](kll_sketch<T>& self) {
66
75
  std::ostringstream oss;
67
76
  self.serialize(oss);
68
77
  return oss.str();
@@ -70,14 +79,14 @@ void bind_kll_sketch(Rice::Module& m, const char* name) {
70
79
  // TODO change to summary?
71
80
  .define_method(
72
81
  "to_string",
73
- *[](datasketches::kll_sketch<T>& self) {
82
+ [](kll_sketch<T>& self) {
74
83
  return self.to_string();
75
84
  })
76
- .define_singleton_method(
85
+ .define_singleton_function(
77
86
  "deserialize",
78
- *[](std::string& is) {
87
+ [](const std::string& is) {
79
88
  std::istringstream iss(is);
80
- return datasketches::kll_sketch<T>::deserialize(iss);
89
+ return kll_sketch<T>::deserialize(iss);
81
90
  });
82
91
  }
83
92