carray 3.0.2 → 3.0.4

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (247) hide show
  1. checksums.yaml +4 -4
  2. data/CHANGELOG.md +568 -0
  3. data/README.md +6 -6
  4. data/carray.gemspec +1 -1
  5. data/ext/ca_axis_descriptor.h +27 -24
  6. data/ext/ca_axis_dispatch.c +220 -0
  7. data/ext/ca_axis_group.c +201 -123
  8. data/ext/ca_binop_dispatch.c +46 -52
  9. data/ext/ca_binop_dispatch.h +1 -1
  10. data/ext/ca_categorical_iterator.c +249 -183
  11. data/ext/ca_compare.c +7 -12
  12. data/ext/ca_for_buffer.h +21 -68
  13. data/ext/ca_group_iter.c +38 -23
  14. data/ext/ca_kernel_iterator.c +1765 -2076
  15. data/ext/ca_kernel_iterator.h +584 -566
  16. data/ext/ca_monop_dispatch.c +86 -63
  17. data/ext/ca_monop_dispatch.h +5 -0
  18. data/ext/ca_obj_array.c +223 -56
  19. data/ext/ca_obj_bincmp.c +172 -138
  20. data/ext/ca_obj_binop.c +116 -230
  21. data/ext/ca_obj_bitarray.c +40 -26
  22. data/ext/ca_obj_bitfield.c +100 -68
  23. data/ext/ca_obj_block.c +19 -9
  24. data/ext/ca_obj_byte_swap.c +45 -46
  25. data/ext/ca_obj_const_string.c +269 -44
  26. data/ext/ca_obj_face.c +612 -89
  27. data/ext/ca_obj_face.h +111 -47
  28. data/ext/ca_obj_fake.c +36 -79
  29. data/ext/ca_obj_farray.c +7 -3
  30. data/ext/ca_obj_field.c +33 -8
  31. data/ext/ca_obj_fixlen_string.c +23 -24
  32. data/ext/ca_obj_grid.c +541 -210
  33. data/ext/ca_obj_meld.c +331 -111
  34. data/ext/ca_obj_moncmp.c +80 -65
  35. data/ext/ca_obj_monop.c +177 -149
  36. data/ext/ca_obj_object.c +664 -248
  37. data/ext/ca_obj_record.c +63 -23
  38. data/ext/ca_obj_reduce.c +18 -16
  39. data/ext/ca_obj_refer.c +140 -58
  40. data/ext/ca_obj_remap.c +85 -24
  41. data/ext/ca_obj_repeat.c +24 -9
  42. data/ext/ca_obj_roll.c +121 -16
  43. data/ext/ca_obj_select.c +61 -58
  44. data/ext/ca_obj_select_axis.c +73 -107
  45. data/ext/ca_obj_shift.c +32 -102
  46. data/ext/ca_obj_stack.c +372 -122
  47. data/ext/ca_obj_stride.c +348 -140
  48. data/ext/ca_obj_string.c +88 -22
  49. data/ext/ca_obj_tile.c +38 -21
  50. data/ext/ca_obj_time.c +45 -31
  51. data/ext/ca_obj_timedelta.c +43 -29
  52. data/ext/ca_obj_transpose.c +17 -6
  53. data/ext/ca_obj_triop.c +125 -141
  54. data/ext/ca_obj_window.c +330 -44
  55. data/ext/ca_op_ipower.c +24 -11
  56. data/ext/ca_op_powi.h +57 -30
  57. data/ext/ca_pow5_table.h +656 -0
  58. data/ext/ca_sort_kernels.h +70 -0
  59. data/ext/ca_sweep_engine.c +104 -81
  60. data/ext/ca_sweep_engine.h +7 -20
  61. data/ext/ca_triop_dispatch.c +4 -1
  62. data/ext/ca_triop_dispatch.h +5 -2
  63. data/ext/caframe_csv_reader.c +542 -0
  64. data/ext/carray.h +241 -92
  65. data/ext/carray_access.c +608 -276
  66. data/ext/carray_address_basis.c +494 -97
  67. data/ext/carray_attribute.c +129 -4
  68. data/ext/carray_bincount.c +39 -7
  69. data/ext/carray_broadcast.c +56 -19
  70. data/ext/carray_build_flags.h +1 -1
  71. data/ext/carray_call_cfunc.c +108 -28
  72. data/ext/carray_cast.c +1576 -223
  73. data/ext/carray_cast_func.rb +77 -24
  74. data/ext/carray_class.c +4 -3
  75. data/ext/carray_conversion.c +208 -75
  76. data/ext/carray_copy.c +82 -24
  77. data/ext/carray_core.c +646 -162
  78. data/ext/carray_count.c +171 -88
  79. data/ext/carray_element.c +94 -60
  80. data/ext/carray_factorize.c +670 -324
  81. data/ext/carray_generate.c +260 -83
  82. data/ext/carray_histogram.c +209 -106
  83. data/ext/carray_hold.c +14 -16
  84. data/ext/carray_index_classifier.c +422 -243
  85. data/ext/carray_internal.h +692 -3
  86. data/ext/carray_kernels_bincmp.c +7 -0
  87. data/ext/carray_kernels_binop.c +487 -120
  88. data/ext/carray_kernels_init.c +653 -50
  89. data/ext/carray_kernels_moncmp.c +15 -8
  90. data/ext/carray_kernels_monop.c +99 -43
  91. data/ext/carray_kernels_reduce_aggregate.c +3925 -6112
  92. data/ext/carray_kernels_reduce_boolean.c +112 -36
  93. data/ext/carray_kernels_reduce_cumulative.c +2127 -6903
  94. data/ext/carray_kernels_reduce_extreme.c +3032 -2892
  95. data/ext/carray_kernels_reduce_variance.c +439 -500
  96. data/ext/carray_kernels_scan.c +1097 -488
  97. data/ext/carray_kernels_search.c +3787 -2176
  98. data/ext/carray_kernels_sort.c +1200 -920
  99. data/ext/carray_kernels_triop.c +479 -7
  100. data/ext/carray_key_windows.c +216 -0
  101. data/ext/carray_lazy.c +967 -153
  102. data/ext/carray_loop.c +81 -65
  103. data/ext/carray_mask.c +910 -324
  104. data/ext/carray_math_kernel.h +1 -0
  105. data/ext/carray_median_percentile.c +299 -100
  106. data/ext/carray_memory_view.c +872 -758
  107. data/ext/carray_operator.c +1171 -364
  108. data/ext/carray_order.c +505 -133
  109. data/ext/carray_partition.c +112 -65
  110. data/ext/carray_random.c +327 -239
  111. data/ext/carray_round.c +273 -0
  112. data/ext/carray_scatter.c +346 -78
  113. data/ext/carray_segment.c +257 -0
  114. data/ext/carray_slab.c +249 -86
  115. data/ext/carray_sort.c +224 -227
  116. data/ext/carray_sort_kernel.c +8 -1
  117. data/ext/carray_struct.c +118 -39
  118. data/ext/carray_test.c +193 -69
  119. data/ext/carray_time_text.c +1182 -0
  120. data/ext/carray_undef.c +4 -0
  121. data/ext/carray_utils.c +579 -113
  122. data/ext/extconf.rb +27 -2
  123. data/ext/mk_call_cfunc.rb +19 -4
  124. data/ext/mk_pow5_table.rb +35 -0
  125. data/ext/mkkernel.rb +1202 -1014
  126. data/ext/ruby_carray.c +17 -5
  127. data/ext/version.h +4 -4
  128. data/ext/version.rb +4 -2
  129. data/lib/carray/attribute.rb +160 -17
  130. data/lib/carray/autoload_carray.rb +23 -27
  131. data/lib/carray/autoload_method_extension.rb +20 -5
  132. data/lib/carray/axis_group.rb +360 -63
  133. data/lib/carray/basics.rb +162 -28
  134. data/lib/carray/bincount_nd.rb +18 -2
  135. data/lib/carray/block_iterator.rb +231 -58
  136. data/lib/carray/boolean_reduce.rb +34 -42
  137. data/lib/carray/categorical.rb +242 -52
  138. data/lib/carray/categorical_iterator.rb +437 -827
  139. data/lib/carray/complex.rb +41 -34
  140. data/lib/carray/conditional.rb +150 -14
  141. data/lib/carray/const_string.rb +59 -4
  142. data/lib/carray/construct.rb +15 -11
  143. data/lib/carray/data_type_extension.rb +1 -0
  144. data/lib/carray/distance_limit.rb +77 -0
  145. data/lib/carray/fixlen_string.rb +7 -1
  146. data/lib/carray/frame/column_verbs.rb +70 -0
  147. data/lib/carray/frame/concat.rb +157 -6
  148. data/lib/carray/frame/convert.rb +82 -7
  149. data/lib/carray/frame/csv_parser.rb +688 -87
  150. data/lib/carray/frame/describe.rb +80 -0
  151. data/lib/carray/frame/frame.rb +369 -45
  152. data/lib/carray/frame/group.rb +631 -90
  153. data/lib/carray/frame/io.rb +506 -79
  154. data/lib/carray/frame/join.rb +411 -62
  155. data/lib/carray/frame/records.rb +116 -19
  156. data/lib/carray/frame/reshape.rb +455 -0
  157. data/lib/carray/frame/sort.rb +30 -18
  158. data/lib/carray/frame/verbs.rb +868 -61
  159. data/lib/carray/frame/windows.rb +117 -0
  160. data/lib/carray/frame.rb +4 -0
  161. data/lib/carray/fuse_source.rb +118 -2
  162. data/lib/carray/fusion.rb +257 -42
  163. data/lib/carray/histogram.rb +6 -1
  164. data/lib/carray/inspect.rb +54 -13
  165. data/lib/carray/iterator.rb +99 -9
  166. data/lib/carray/key_windows.rb +620 -0
  167. data/lib/carray/lazy.rb +216 -193
  168. data/lib/carray/mask_gap_fill.rb +112 -92
  169. data/lib/carray/math.rb +32 -19
  170. data/lib/carray/meld_reduce.rb +52 -21
  171. data/lib/carray/methods/align_addr.rb +8 -4
  172. data/lib/carray/methods/bin.rb +26 -18
  173. data/lib/carray/methods/bincount.rb +18 -2
  174. data/lib/carray/methods/bit_string.rb +15 -4
  175. data/lib/carray/methods/choose.rb +16 -3
  176. data/lib/carray/methods/composition.rb +18 -10
  177. data/lib/carray/methods/factorize.rb +2 -2
  178. data/lib/carray/methods/gather_nd.rb +25 -5
  179. data/lib/carray/methods/insert_block.rb +9 -14
  180. data/lib/carray/methods/is_between.rb +59 -0
  181. data/lib/carray/methods/is_in.rb +140 -10
  182. data/lib/carray/methods/join.rb +2 -7
  183. data/lib/carray/methods/lag.rb +49 -0
  184. data/lib/carray/methods/locate_addr.rb +182 -45
  185. data/lib/carray/methods/mode.rb +18 -1
  186. data/lib/carray/methods/nunique.rb +11 -6
  187. data/lib/carray/methods/repeat.rb +1 -20
  188. data/lib/carray/methods/resize.rb +21 -23
  189. data/lib/carray/methods/snap.rb +34 -12
  190. data/lib/carray/methods/unique.rb +11 -2
  191. data/lib/carray/methods/value_counts.rb +4 -11
  192. data/lib/carray/record_order.rb +208 -0
  193. data/lib/carray/rng.rb +3 -0
  194. data/lib/carray/runtime.rb +65 -17
  195. data/lib/carray/segment_iterator.rb +1221 -0
  196. data/lib/carray/serialize.rb +214 -30
  197. data/lib/carray/slab_iterator.rb +101 -27
  198. data/lib/carray/stack.rb +10 -4
  199. data/lib/carray/string.rb +105 -5
  200. data/lib/carray/string_operation_extension.rb +12 -4
  201. data/lib/carray/struct.rb +137 -58
  202. data/lib/carray/struct_builder.rb +124 -17
  203. data/lib/carray/time.rb +1222 -186
  204. data/lib/carray/window_iterator.rb +242 -81
  205. data/lib/carray.rb +9 -2
  206. data/yard-stubs/ca_obj_array.rb +4 -4
  207. data/yard-stubs/ca_obj_bitarray.rb +7 -0
  208. data/yard-stubs/ca_obj_bitfield.rb +23 -12
  209. data/yard-stubs/ca_obj_fake.rb +1 -1
  210. data/yard-stubs/ca_obj_field.rb +5 -1
  211. data/yard-stubs/ca_obj_grid.rb +21 -10
  212. data/yard-stubs/ca_obj_refer.rb +5 -4
  213. data/yard-stubs/ca_obj_roll.rb +2 -3
  214. data/yard-stubs/ca_obj_shift.rb +5 -1
  215. data/yard-stubs/ca_obj_stride.rb +10 -7
  216. data/yard-stubs/ca_obj_tile.rb +1 -1
  217. data/yard-stubs/ca_obj_transpose.rb +1 -1
  218. data/yard-stubs/ca_obj_window.rb +9 -5
  219. data/yard-stubs/carray_access.rb +11 -5
  220. data/yard-stubs/carray_attribute.rb +0 -5
  221. data/yard-stubs/carray_broadcast.rb +1 -1
  222. data/yard-stubs/carray_cast.rb +88 -43
  223. data/yard-stubs/carray_conversion.rb +15 -2
  224. data/yard-stubs/carray_copy.rb +12 -5
  225. data/yard-stubs/carray_core.rb +3 -3
  226. data/yard-stubs/carray_count.rb +9 -0
  227. data/yard-stubs/carray_element.rb +10 -4
  228. data/yard-stubs/carray_generate.rb +21 -1
  229. data/yard-stubs/carray_lazy.rb +5 -1
  230. data/yard-stubs/carray_loop.rb +4 -1
  231. data/yard-stubs/carray_mask.rb +60 -12
  232. data/yard-stubs/carray_math.rb +52 -14
  233. data/yard-stubs/carray_median_percentile.rb +15 -8
  234. data/yard-stubs/carray_memory_view.rb +40 -23
  235. data/yard-stubs/carray_order.rb +61 -29
  236. data/yard-stubs/carray_random.rb +13 -4
  237. data/yard-stubs/carray_scatter.rb +21 -9
  238. data/yard-stubs/carray_segment.rb +72 -0
  239. data/yard-stubs/carray_slab.rb +10 -2
  240. data/yard-stubs/carray_sort.rb +49 -8
  241. data/yard-stubs/carray_test.rb +15 -3
  242. data/yard-stubs/carray_utils.rb +0 -26
  243. metadata +19 -5
  244. data/ext/ca_for_each_element.h +0 -239
  245. data/ext/carray_kernels_map.c +0 -3467
  246. data/lib/carray/arrow_tensor.rb +0 -401
  247. data/lib/carray/table.rb +0 -108
@@ -38,12 +38,40 @@ class CAFrame
38
38
  def sort_by_key(*specs, order: :asc, masked_position: :last)
39
39
  raise ArgumentError, "sort_by_key requires at least one key" if specs.empty?
40
40
  validate_sort_order(order)
41
- keys = specs.map do |spec|
41
+ named = specs.map do |spec|
42
42
  name, dir = parse_sort_key_spec(spec, order)
43
- col = sort_key_column(name)
43
+ [column_key(name, "sort_by_key"), key_column(column_key(name, "sort_by_key"), "sort_by_key"), dir]
44
+ end
45
+ CAFrame.__send__(:refuse_repeated_names, named.map(&:first), "sort_by_key")
46
+ keys = named.map do |name, col, dir|
44
47
  dir == :desc ? col.order(descending: true, method: :dense) : col
45
48
  end
46
49
  select_rows(CArray.sort_addr(*keys, masked_position: masked_position))
50
+ rescue ArgumentError, CArray::DataTypeError => e
51
+ raise unless named
52
+ name, col, = named.find { |_, c, _| !sortable_key?(c) }
53
+ raise unless name
54
+ reason = e.message.sub(/\A\w+: /, "")
55
+ if col.face?
56
+ # The core names the conversion as an array method; in a frame it is
57
+ # written with sort_by.
58
+ reason = "#{reason.split(";").first}; sort by its values with " \
59
+ "sort_by { |f| f[#{name.inspect}].to_type(:object) }"
60
+ if (storage = col.__send__(:face_storage_method))
61
+ reason += ", or by its storage with " \
62
+ "sort_by { |f| f[#{name.inspect}].#{storage} }"
63
+ end
64
+ end
65
+ raise e.class, "sort_by_key: the column #{name.inspect} cannot be sorted: #{reason}"
66
+ end
67
+
68
+ # Whether +col+ can be a sort key, asked of an empty slice of it so that
69
+ # nothing is sorted: used only to name the column a refusal was about.
70
+ private def sortable_key?(col)
71
+ CArray.sort_addr(col[0...0])
72
+ true
73
+ rescue ArgumentError, CArray::DataTypeError
74
+ false
47
75
  end
48
76
 
49
77
  # Sort rows by key columns the block builds (memo §4.3 escape). The block
@@ -94,20 +122,4 @@ class CAFrame
94
122
  raise ArgumentError, "sort order must be :asc or :desc, got #{dir.inspect}"
95
123
  end
96
124
 
97
- private def sort_key_column(name)
98
- key = name.to_s
99
- col = if @columns.key?(key)
100
- @columns[key]
101
- elsif @index && @axis_name == key
102
- @index
103
- else
104
- raise KeyError, "no column #{key.inspect}"
105
- end
106
- unless col.ndim == 1
107
- raise ArgumentError,
108
- "sort_by_key: key column #{key.inspect} is #{col.ndim}-D; " \
109
- "sort by a scalar (1-D) column"
110
- end
111
- col
112
- end
113
125
  end