unit_graph.cc 41.9 KB
Newer Older
1
2
/*!
 *  Copyright (c) 2019 by Contributors
Minjie Wang's avatar
Minjie Wang committed
3
4
 * \file graph/unit_graph.cc
 * \brief UnitGraph graph implementation
5
6
7
8
 */
#include <dgl/array.h>
#include <dgl/lazy.h>
#include <dgl/immutable_graph.h>
9
#include <dgl/base_heterograph.h>
10

Minjie Wang's avatar
Minjie Wang committed
11
#include "./unit_graph.h"
12
13
14
#include "../c_api_common.h"

namespace dgl {
15

16
namespace {
17
18
19

using namespace dgl::aten;

Minjie Wang's avatar
Minjie Wang committed
20
21
22
23
24
25
26
27
28
29
// create metagraph of one node type
inline GraphPtr CreateUnitGraphMetaGraph1() {
  // a self-loop edge 0->0
  std::vector<int64_t> row_vec(1, 0);
  std::vector<int64_t> col_vec(1, 0);
  IdArray row = aten::VecToIdArray(row_vec);
  IdArray col = aten::VecToIdArray(col_vec);
  GraphPtr g = ImmutableGraph::CreateFromCOO(1, row, col);
  return g;
}
30

Minjie Wang's avatar
Minjie Wang committed
31
32
33
34
35
// create metagraph of two node types
inline GraphPtr CreateUnitGraphMetaGraph2() {
  // an edge 0->1
  std::vector<int64_t> row_vec(1, 0);
  std::vector<int64_t> col_vec(1, 1);
36
37
38
39
40
  IdArray row = aten::VecToIdArray(row_vec);
  IdArray col = aten::VecToIdArray(col_vec);
  GraphPtr g = ImmutableGraph::CreateFromCOO(2, row, col);
  return g;
}
Minjie Wang's avatar
Minjie Wang committed
41
42
43
44
45
46
47
48
49
50
51
52

inline GraphPtr CreateUnitGraphMetaGraph(int num_vtypes) {
  static GraphPtr mg1 = CreateUnitGraphMetaGraph1();
  static GraphPtr mg2 = CreateUnitGraphMetaGraph2();
  if (num_vtypes == 1)
    return mg1;
  else if (num_vtypes == 2)
    return mg2;
  else
    LOG(FATAL) << "Invalid number of vertex types. Must be 1 or 2.";
  return {};
}
53
54

};  // namespace
55
56
57
58
59
60
61

//////////////////////////////////////////////////////////
//
// COO graph implementation
//
//////////////////////////////////////////////////////////

Minjie Wang's avatar
Minjie Wang committed
62
class UnitGraph::COO : public BaseHeteroGraph {
63
 public:
Minjie Wang's avatar
Minjie Wang committed
64
65
  COO(GraphPtr metagraph, int64_t num_src, int64_t num_dst, IdArray src, IdArray dst)
    : BaseHeteroGraph(metagraph) {
66
67
68
    CHECK(aten::IsValidIdArray(src));
    CHECK(aten::IsValidIdArray(dst));
    CHECK_EQ(src->shape[0], dst->shape[0]) << "Input arrays should have the same length.";
69
70
    adj_ = aten::COOMatrix{num_src, num_dst, src, dst};
  }
71

Minjie Wang's avatar
Minjie Wang committed
72
73
74
  COO(GraphPtr metagraph, int64_t num_src, int64_t num_dst,
      IdArray src, IdArray dst, bool is_multigraph)
    : BaseHeteroGraph(metagraph),
75
      is_multigraph_(is_multigraph) {
76
77
78
    CHECK(aten::IsValidIdArray(src));
    CHECK(aten::IsValidIdArray(dst));
    CHECK_EQ(src->shape[0], dst->shape[0]) << "Input arrays should have the same length.";
79
80
    adj_ = aten::COOMatrix{num_src, num_dst, src, dst};
  }
81

82
83
84
85
  COO(GraphPtr metagraph, const aten::COOMatrix& coo)
    : BaseHeteroGraph(metagraph), adj_(coo) {
    // Data index should not be inherited. Edges in COO format are always
    // assigned ids from 0 to num_edges - 1.
86
    adj_.data = aten::NullArray();
87
  }
88

Minjie Wang's avatar
Minjie Wang committed
89
90
  inline dgl_type_t SrcType() const {
    return 0;
91
  }
Minjie Wang's avatar
Minjie Wang committed
92
93
94
95
96
97
98

  inline dgl_type_t DstType() const {
    return NumVertexTypes() == 1? 0 : 1;
  }

  inline dgl_type_t EdgeType() const {
    return 0;
99
100
101
  }

  HeteroGraphPtr GetRelationGraph(dgl_type_t etype) const override {
Minjie Wang's avatar
Minjie Wang committed
102
    LOG(FATAL) << "The method shouldn't be called for UnitGraph graph. "
103
104
105
106
107
      << "The relation graph is simply this graph itself.";
    return {};
  }

  void AddVertices(dgl_type_t vtype, uint64_t num_vertices) override {
Minjie Wang's avatar
Minjie Wang committed
108
    LOG(FATAL) << "UnitGraph graph is not mutable.";
109
110
111
  }

  void AddEdge(dgl_type_t etype, dgl_id_t src, dgl_id_t dst) override {
Minjie Wang's avatar
Minjie Wang committed
112
    LOG(FATAL) << "UnitGraph graph is not mutable.";
113
114
115
  }

  void AddEdges(dgl_type_t etype, IdArray src_ids, IdArray dst_ids) override {
Minjie Wang's avatar
Minjie Wang committed
116
    LOG(FATAL) << "UnitGraph graph is not mutable.";
117
118
119
  }

  void Clear() override {
Minjie Wang's avatar
Minjie Wang committed
120
    LOG(FATAL) << "UnitGraph graph is not mutable.";
121
122
  }

123
124
125
126
  DLDataType DataType() const override {
    return adj_.row->dtype;
  }

127
128
129
130
131
132
133
134
  DLContext Context() const override {
    return adj_.row->ctx;
  }

  uint8_t NumBits() const override {
    return adj_.row->dtype.bits;
  }

135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
  COO AsNumBits(uint8_t bits) const {
    if (NumBits() == bits)
      return *this;

    COO ret(
        meta_graph_,
        adj_.num_rows, adj_.num_cols,
        aten::AsNumBits(adj_.row, bits),
        aten::AsNumBits(adj_.col, bits));
    ret.is_multigraph_ = is_multigraph_;
    return ret;
  }

  COO CopyTo(const DLContext& ctx) const {
    if (Context() == ctx)
      return *this;

    COO ret(
        meta_graph_,
        adj_.num_rows, adj_.num_cols,
        adj_.row.CopyTo(ctx),
        adj_.col.CopyTo(ctx));
    ret.is_multigraph_ = is_multigraph_;
    return ret;
  }

161
162
163
164
165
166
167
168
169
170
171
  bool IsMultigraph() const override {
    return const_cast<COO*>(this)->is_multigraph_.Get([this] () {
        return aten::COOHasDuplicate(adj_);
      });
  }

  bool IsReadonly() const override {
    return true;
  }

  uint64_t NumVertices(dgl_type_t vtype) const override {
Minjie Wang's avatar
Minjie Wang committed
172
    if (vtype == SrcType()) {
173
      return adj_.num_rows;
Minjie Wang's avatar
Minjie Wang committed
174
    } else if (vtype == DstType()) {
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
      return adj_.num_cols;
    } else {
      LOG(FATAL) << "Invalid vertex type: " << vtype;
      return 0;
    }
  }

  uint64_t NumEdges(dgl_type_t etype) const override {
    return adj_.row->shape[0];
  }

  bool HasVertex(dgl_type_t vtype, dgl_id_t vid) const override {
    return vid < NumVertices(vtype);
  }

  BoolArray HasVertices(dgl_type_t vtype, IdArray vids) const override {
    LOG(FATAL) << "Not enabled for COO graph";
    return {};
  }

  bool HasEdgeBetween(dgl_type_t etype, dgl_id_t src, dgl_id_t dst) const override {
196
197
198
    CHECK(HasVertex(SrcType(), src)) << "Invalid src vertex id: " << src;
    CHECK(HasVertex(DstType(), dst)) << "Invalid dst vertex id: " << dst;
    return aten::COOIsNonZero(adj_, src, dst);
199
200
201
  }

  BoolArray HasEdgesBetween(dgl_type_t etype, IdArray src_ids, IdArray dst_ids) const override {
202
203
204
    CHECK(aten::IsValidIdArray(src_ids)) << "Invalid vertex id array.";
    CHECK(aten::IsValidIdArray(dst_ids)) << "Invalid vertex id array.";
    return aten::COOIsNonZero(adj_, src_ids, dst_ids);
205
206
207
  }

  IdArray Predecessors(dgl_type_t etype, dgl_id_t dst) const override {
208
209
    CHECK(HasVertex(DstType(), dst)) << "Invalid dst vertex id: " << dst;
    return aten::COOGetRowDataAndIndices(aten::COOTranspose(adj_), dst).second;
210
211
212
  }

  IdArray Successors(dgl_type_t etype, dgl_id_t src) const override {
213
214
    CHECK(HasVertex(SrcType(), src)) << "Invalid src vertex id: " << src;
    return aten::COOGetRowDataAndIndices(adj_, src).second;
215
216
217
  }

  IdArray EdgeId(dgl_type_t etype, dgl_id_t src, dgl_id_t dst) const override {
218
219
220
    CHECK(HasVertex(SrcType(), src)) << "Invalid src vertex id: " << src;
    CHECK(HasVertex(DstType(), dst)) << "Invalid dst vertex id: " << dst;
    return aten::COOGetData(adj_, src, dst);
221
222
223
  }

  EdgeArray EdgeIds(dgl_type_t etype, IdArray src, IdArray dst) const override {
224
225
226
227
    CHECK(aten::IsValidIdArray(src)) << "Invalid vertex id array.";
    CHECK(aten::IsValidIdArray(dst)) << "Invalid vertex id array.";
    const auto& arrs = aten::COOGetDataAndIndices(adj_, src, dst);
    return EdgeArray{arrs[0], arrs[1], arrs[2]};
228
229
230
231
  }

  std::pair<dgl_id_t, dgl_id_t> FindEdge(dgl_type_t etype, dgl_id_t eid) const override {
    CHECK(eid < NumEdges(etype)) << "Invalid edge id: " << eid;
232
233
    const dgl_id_t src = aten::IndexSelect<int64_t>(adj_.row, eid);
    const dgl_id_t dst = aten::IndexSelect<int64_t>(adj_.col, eid);
234
235
236
237
    return std::pair<dgl_id_t, dgl_id_t>(src, dst);
  }

  EdgeArray FindEdges(dgl_type_t etype, IdArray eids) const override {
238
    CHECK(aten::IsValidIdArray(eids)) << "Invalid edge id array";
239
240
241
242
243
244
    return EdgeArray{aten::IndexSelect(adj_.row, eids),
                     aten::IndexSelect(adj_.col, eids),
                     eids};
  }

  EdgeArray InEdges(dgl_type_t etype, dgl_id_t vid) const override {
245
246
247
248
249
    IdArray ret_src, ret_eid;
    std::tie(ret_eid, ret_src) = aten::COOGetRowDataAndIndices(
        aten::COOTranspose(adj_), vid);
    IdArray ret_dst = aten::Full(vid, ret_src->shape[0], NumBits(), ret_src->ctx);
    return EdgeArray{ret_src, ret_dst, ret_eid};
250
251
252
  }

  EdgeArray InEdges(dgl_type_t etype, IdArray vids) const override {
253
254
255
256
    CHECK(aten::IsValidIdArray(vids)) << "Invalid vertex id array.";
    auto coosubmat = aten::COOSliceRows(aten::COOTranspose(adj_), vids);
    auto row = aten::IndexSelect(vids, coosubmat.row);
    return EdgeArray{coosubmat.col, row, coosubmat.data};
257
258
259
  }

  EdgeArray OutEdges(dgl_type_t etype, dgl_id_t vid) const override {
260
261
262
263
    IdArray ret_dst, ret_eid;
    std::tie(ret_eid, ret_dst) = aten::COOGetRowDataAndIndices(adj_, vid);
    IdArray ret_src = aten::Full(vid, ret_dst->shape[0], NumBits(), ret_dst->ctx);
    return EdgeArray{ret_src, ret_dst, ret_eid};
264
265
266
  }

  EdgeArray OutEdges(dgl_type_t etype, IdArray vids) const override {
267
268
269
270
    CHECK(aten::IsValidIdArray(vids)) << "Invalid vertex id array.";
    auto coosubmat = aten::COOSliceRows(adj_, vids);
    auto row = aten::IndexSelect(vids, coosubmat.row);
    return EdgeArray{row, coosubmat.col, coosubmat.data};
271
272
273
274
275
276
277
278
279
280
281
  }

  EdgeArray Edges(dgl_type_t etype, const std::string &order = "") const override {
    CHECK(order.empty() || order == std::string("eid"))
      << "COO only support Edges of order \"eid\", but got \""
      << order << "\".";
    IdArray rst_eid = aten::Range(0, NumEdges(etype), NumBits(), Context());
    return EdgeArray{adj_.row, adj_.col, rst_eid};
  }

  uint64_t InDegree(dgl_type_t etype, dgl_id_t vid) const override {
282
283
    CHECK(HasVertex(DstType(), vid)) << "Invalid dst vertex id: " << vid;
    return aten::COOGetRowNNZ(aten::COOTranspose(adj_), vid);
284
285
286
  }

  DegreeArray InDegrees(dgl_type_t etype, IdArray vids) const override {
287
288
    CHECK(aten::IsValidIdArray(vids)) << "Invalid vertex id array.";
    return aten::COOGetRowNNZ(aten::COOTranspose(adj_), vids);
289
290
291
  }

  uint64_t OutDegree(dgl_type_t etype, dgl_id_t vid) const override {
292
293
    CHECK(HasVertex(SrcType(), vid)) << "Invalid src vertex id: " << vid;
    return aten::COOGetRowNNZ(adj_, vid);
294
295
296
  }

  DegreeArray OutDegrees(dgl_type_t etype, IdArray vids) const override {
297
298
    CHECK(aten::IsValidIdArray(vids)) << "Invalid vertex id array.";
    return aten::COOGetRowNNZ(adj_, vids);
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
  }

  DGLIdIters SuccVec(dgl_type_t etype, dgl_id_t vid) const override {
    LOG(INFO) << "Not enabled for COO graph.";
    return {};
  }

  DGLIdIters OutEdgeVec(dgl_type_t etype, dgl_id_t vid) const override {
    LOG(INFO) << "Not enabled for COO graph.";
    return {};
  }

  DGLIdIters PredVec(dgl_type_t etype, dgl_id_t vid) const override {
    LOG(INFO) << "Not enabled for COO graph.";
    return {};
  }

  DGLIdIters InEdgeVec(dgl_type_t etype, dgl_id_t vid) const override {
    LOG(INFO) << "Not enabled for COO graph.";
    return {};
  }

  std::vector<IdArray> GetAdj(
      dgl_type_t etype, bool transpose, const std::string &fmt) const override {
    CHECK(fmt == "coo") << "Not valid adj format request.";
    if (transpose) {
      return {aten::HStack(adj_.col, adj_.row)};
    } else {
      return {aten::HStack(adj_.row, adj_.col)};
    }
  }

331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
  aten::COOMatrix GetCOOMatrix(dgl_type_t etype) const override {
    return adj_;
  }

  aten::CSRMatrix GetCSCMatrix(dgl_type_t etype) const override {
    LOG(FATAL) << "Not enabled for COO graph";
    return aten::CSRMatrix();
  }

  aten::CSRMatrix GetCSRMatrix(dgl_type_t etype) const override {
    LOG(FATAL) << "Not enabled for COO graph";
    return aten::CSRMatrix();
  }

  SparseFormat SelectFormat(dgl_type_t etype, SparseFormat preferred_format) const override {
    LOG(FATAL) << "Not enabled for COO graph";
    return SparseFormat::ANY;
  }

350
  HeteroSubgraph VertexSubgraph(const std::vector<IdArray>& vids) const override {
351
352
353
354
355
356
357
358
359
360
361
362
    CHECK_EQ(vids.size(), NumVertexTypes()) << "Number of vertex types mismatch";
    auto srcvids = vids[SrcType()], dstvids = vids[DstType()];
    CHECK(aten::IsValidIdArray(srcvids)) << "Invalid vertex id array.";
    CHECK(aten::IsValidIdArray(dstvids)) << "Invalid vertex id array.";
    HeteroSubgraph subg;
    const auto& submat = aten::COOSliceMatrix(adj_, srcvids, dstvids);
    IdArray sub_eids = aten::Range(0, submat.data->shape[0], NumBits(), Context());
    subg.graph = std::make_shared<COO>(meta_graph(), submat.num_rows, submat.num_cols,
        submat.row, submat.col);
    subg.induced_vertices = vids;
    subg.induced_edges.emplace_back(submat.data);
    return subg;
363
364
365
366
367
368
369
370
371
372
373
374
375
376
  }

  HeteroSubgraph EdgeSubgraph(
      const std::vector<IdArray>& eids, bool preserve_nodes = false) const override {
    CHECK_EQ(eids.size(), 1) << "Edge type number mismatch.";
    HeteroSubgraph subg;
    if (!preserve_nodes) {
      IdArray new_src = aten::IndexSelect(adj_.row, eids[0]);
      IdArray new_dst = aten::IndexSelect(adj_.col, eids[0]);
      subg.induced_vertices.emplace_back(aten::Relabel_({new_src}));
      subg.induced_vertices.emplace_back(aten::Relabel_({new_dst}));
      const auto new_nsrc = subg.induced_vertices[0]->shape[0];
      const auto new_ndst = subg.induced_vertices[1]->shape[0];
      subg.graph = std::make_shared<COO>(
Minjie Wang's avatar
Minjie Wang committed
377
          meta_graph(), new_nsrc, new_ndst, new_src, new_dst);
378
379
380
381
382
383
384
      subg.induced_edges = eids;
    } else {
      IdArray new_src = aten::IndexSelect(adj_.row, eids[0]);
      IdArray new_dst = aten::IndexSelect(adj_.col, eids[0]);
      subg.induced_vertices.emplace_back(aten::Range(0, NumVertices(0), NumBits(), Context()));
      subg.induced_vertices.emplace_back(aten::Range(0, NumVertices(1), NumBits(), Context()));
      subg.graph = std::make_shared<COO>(
Minjie Wang's avatar
Minjie Wang committed
385
          meta_graph(), NumVertices(0), NumVertices(1), new_src, new_dst);
386
387
388
389
390
391
392
393
394
      subg.induced_edges = eids;
    }
    return subg;
  }

  aten::COOMatrix adj() const {
    return adj_;
  }

395
396
397
398
399
400
401
402
403
  /*!
   * \brief Determines whether the graph is "hypersparse", i.e. having significantly more
   * nodes than edges.
   */
  bool IsHypersparse() const {
    return (NumVertices(SrcType()) / 8 > NumEdges(EdgeType())) &&
           (NumVertices(SrcType()) > 1000000);
  }

404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
 private:
  /*! \brief internal adjacency matrix. Data array is empty */
  aten::COOMatrix adj_;

  /*! \brief multi-graph flag */
  Lazy<bool> is_multigraph_;
};

//////////////////////////////////////////////////////////
//
// CSR graph implementation
//
//////////////////////////////////////////////////////////

/*! \brief CSR graph */
Minjie Wang's avatar
Minjie Wang committed
419
class UnitGraph::CSR : public BaseHeteroGraph {
420
 public:
Minjie Wang's avatar
Minjie Wang committed
421
  CSR(GraphPtr metagraph, int64_t num_src, int64_t num_dst,
422
      IdArray indptr, IdArray indices, IdArray edge_ids)
Minjie Wang's avatar
Minjie Wang committed
423
    : BaseHeteroGraph(metagraph) {
424
425
426
427
428
    CHECK(aten::IsValidIdArray(indptr));
    CHECK(aten::IsValidIdArray(indices));
    CHECK(aten::IsValidIdArray(edge_ids));
    CHECK_EQ(indices->shape[0], edge_ids->shape[0])
      << "indices and edge id arrays should have the same length";
429
    adj_ = aten::CSRMatrix{num_src, num_dst, indptr, indices, edge_ids};
Da Zheng's avatar
Da Zheng committed
430
    sorted_ = false;
431
432
  }

Minjie Wang's avatar
Minjie Wang committed
433
  CSR(GraphPtr metagraph, int64_t num_src, int64_t num_dst,
434
      IdArray indptr, IdArray indices, IdArray edge_ids, bool is_multigraph)
Minjie Wang's avatar
Minjie Wang committed
435
    : BaseHeteroGraph(metagraph), is_multigraph_(is_multigraph) {
436
437
438
439
440
    CHECK(aten::IsValidIdArray(indptr));
    CHECK(aten::IsValidIdArray(indices));
    CHECK(aten::IsValidIdArray(edge_ids));
    CHECK_EQ(indices->shape[0], edge_ids->shape[0])
      << "indices and edge id arrays should have the same length";
441
    adj_ = aten::CSRMatrix{num_src, num_dst, indptr, indices, edge_ids};
Da Zheng's avatar
Da Zheng committed
442
    sorted_ = false;
443
444
  }

445
  CSR(GraphPtr metagraph, const aten::CSRMatrix& csr)
Da Zheng's avatar
Da Zheng committed
446
447
448
    : BaseHeteroGraph(metagraph), adj_(csr) {
    sorted_ = false;
  }
449

Minjie Wang's avatar
Minjie Wang committed
450
451
  inline dgl_type_t SrcType() const {
    return 0;
452
  }
Minjie Wang's avatar
Minjie Wang committed
453
454
455
456
457
458
459

  inline dgl_type_t DstType() const {
    return NumVertexTypes() == 1? 0 : 1;
  }

  inline dgl_type_t EdgeType() const {
    return 0;
460
461
462
  }

  HeteroGraphPtr GetRelationGraph(dgl_type_t etype) const override {
Minjie Wang's avatar
Minjie Wang committed
463
    LOG(FATAL) << "The method shouldn't be called for UnitGraph graph. "
464
465
466
467
468
      << "The relation graph is simply this graph itself.";
    return {};
  }

  void AddVertices(dgl_type_t vtype, uint64_t num_vertices) override {
Minjie Wang's avatar
Minjie Wang committed
469
    LOG(FATAL) << "UnitGraph graph is not mutable.";
470
471
472
  }

  void AddEdge(dgl_type_t etype, dgl_id_t src, dgl_id_t dst) override {
Minjie Wang's avatar
Minjie Wang committed
473
    LOG(FATAL) << "UnitGraph graph is not mutable.";
474
475
476
  }

  void AddEdges(dgl_type_t etype, IdArray src_ids, IdArray dst_ids) override {
Minjie Wang's avatar
Minjie Wang committed
477
    LOG(FATAL) << "UnitGraph graph is not mutable.";
478
479
480
  }

  void Clear() override {
Minjie Wang's avatar
Minjie Wang committed
481
    LOG(FATAL) << "UnitGraph graph is not mutable.";
482
483
  }

484
485
486
487
  DLDataType DataType() const override {
    return adj_.indices->dtype;
  }

488
489
490
491
492
493
494
495
  DLContext Context() const override {
    return adj_.indices->ctx;
  }

  uint8_t NumBits() const override {
    return adj_.indices->dtype.bits;
  }

496
497
498
499
500
  CSR AsNumBits(uint8_t bits) const {
    if (NumBits() == bits) {
      return *this;
    } else {
      CSR ret(
Minjie Wang's avatar
Minjie Wang committed
501
          meta_graph_,
502
503
504
505
506
507
508
509
510
511
512
513
514
515
          adj_.num_rows, adj_.num_cols,
          aten::AsNumBits(adj_.indptr, bits),
          aten::AsNumBits(adj_.indices, bits),
          aten::AsNumBits(adj_.data, bits));
      ret.is_multigraph_ = is_multigraph_;
      return ret;
    }
  }

  CSR CopyTo(const DLContext& ctx) const {
    if (Context() == ctx) {
      return *this;
    } else {
      CSR ret(
Minjie Wang's avatar
Minjie Wang committed
516
          meta_graph_,
517
518
519
520
521
522
523
524
525
          adj_.num_rows, adj_.num_cols,
          adj_.indptr.CopyTo(ctx),
          adj_.indices.CopyTo(ctx),
          adj_.data.CopyTo(ctx));
      ret.is_multigraph_ = is_multigraph_;
      return ret;
    }
  }

526
527
528
529
530
531
532
533
534
535
536
  bool IsMultigraph() const override {
    return const_cast<CSR*>(this)->is_multigraph_.Get([this] () {
        return aten::CSRHasDuplicate(adj_);
      });
  }

  bool IsReadonly() const override {
    return true;
  }

  uint64_t NumVertices(dgl_type_t vtype) const override {
Minjie Wang's avatar
Minjie Wang committed
537
    if (vtype == SrcType()) {
538
      return adj_.num_rows;
Minjie Wang's avatar
Minjie Wang committed
539
    } else if (vtype == DstType()) {
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
      return adj_.num_cols;
    } else {
      LOG(FATAL) << "Invalid vertex type: " << vtype;
      return 0;
    }
  }

  uint64_t NumEdges(dgl_type_t etype) const override {
    return adj_.indices->shape[0];
  }

  bool HasVertex(dgl_type_t vtype, dgl_id_t vid) const override {
    return vid < NumVertices(vtype);
  }

  BoolArray HasVertices(dgl_type_t vtype, IdArray vids) const override {
    LOG(FATAL) << "Not enabled for COO graph";
    return {};
  }

  bool HasEdgeBetween(dgl_type_t etype, dgl_id_t src, dgl_id_t dst) const override {
Minjie Wang's avatar
Minjie Wang committed
561
562
    CHECK(HasVertex(SrcType(), src)) << "Invalid src vertex id: " << src;
    CHECK(HasVertex(DstType(), dst)) << "Invalid dst vertex id: " << dst;
563
564
565
566
    return aten::CSRIsNonZero(adj_, src, dst);
  }

  BoolArray HasEdgesBetween(dgl_type_t etype, IdArray src_ids, IdArray dst_ids) const override {
567
568
    CHECK(aten::IsValidIdArray(src_ids)) << "Invalid vertex id array.";
    CHECK(aten::IsValidIdArray(dst_ids)) << "Invalid vertex id array.";
569
570
571
572
573
574
575
576
577
    return aten::CSRIsNonZero(adj_, src_ids, dst_ids);
  }

  IdArray Predecessors(dgl_type_t etype, dgl_id_t dst) const override {
    LOG(INFO) << "Not enabled for CSR graph.";
    return {};
  }

  IdArray Successors(dgl_type_t etype, dgl_id_t src) const override {
Minjie Wang's avatar
Minjie Wang committed
578
    CHECK(HasVertex(SrcType(), src)) << "Invalid src vertex id: " << src;
579
580
581
582
    return aten::CSRGetRowColumnIndices(adj_, src);
  }

  IdArray EdgeId(dgl_type_t etype, dgl_id_t src, dgl_id_t dst) const override {
Minjie Wang's avatar
Minjie Wang committed
583
584
    CHECK(HasVertex(SrcType(), src)) << "Invalid src vertex id: " << src;
    CHECK(HasVertex(DstType(), dst)) << "Invalid dst vertex id: " << dst;
585
586
587
588
    return aten::CSRGetData(adj_, src, dst);
  }

  EdgeArray EdgeIds(dgl_type_t etype, IdArray src, IdArray dst) const override {
589
590
    CHECK(aten::IsValidIdArray(src)) << "Invalid vertex id array.";
    CHECK(aten::IsValidIdArray(dst)) << "Invalid vertex id array.";
591
592
593
594
595
    const auto& arrs = aten::CSRGetDataAndIndices(adj_, src, dst);
    return EdgeArray{arrs[0], arrs[1], arrs[2]};
  }

  std::pair<dgl_id_t, dgl_id_t> FindEdge(dgl_type_t etype, dgl_id_t eid) const override {
596
    LOG(FATAL) << "Not enabled for CSR graph.";
597
598
599
600
    return {};
  }

  EdgeArray FindEdges(dgl_type_t etype, IdArray eids) const override {
601
    LOG(FATAL) << "Not enabled for CSR graph.";
602
603
604
605
    return {};
  }

  EdgeArray InEdges(dgl_type_t etype, dgl_id_t vid) const override {
606
    LOG(FATAL) << "Not enabled for CSR graph.";
607
608
609
610
    return {};
  }

  EdgeArray InEdges(dgl_type_t etype, IdArray vids) const override {
611
    LOG(FATAL) << "Not enabled for CSR graph.";
612
613
614
615
    return {};
  }

  EdgeArray OutEdges(dgl_type_t etype, dgl_id_t vid) const override {
Minjie Wang's avatar
Minjie Wang committed
616
    CHECK(HasVertex(SrcType(), vid)) << "Invalid src vertex id: " << vid;
617
618
619
620
621
622
623
    IdArray ret_dst = aten::CSRGetRowColumnIndices(adj_, vid);
    IdArray ret_eid = aten::CSRGetRowData(adj_, vid);
    IdArray ret_src = aten::Full(vid, ret_dst->shape[0], NumBits(), ret_dst->ctx);
    return EdgeArray{ret_src, ret_dst, ret_eid};
  }

  EdgeArray OutEdges(dgl_type_t etype, IdArray vids) const override {
624
    CHECK(aten::IsValidIdArray(vids)) << "Invalid vertex id array.";
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
    auto csrsubmat = aten::CSRSliceRows(adj_, vids);
    auto coosubmat = aten::CSRToCOO(csrsubmat, false);
    // Note that the row id in the csr submat is relabled, so
    // we need to recover it using an index select.
    auto row = aten::IndexSelect(vids, coosubmat.row);
    return EdgeArray{row, coosubmat.col, coosubmat.data};
  }

  EdgeArray Edges(dgl_type_t etype, const std::string &order = "") const override {
    CHECK(order.empty() || order == std::string("srcdst"))
      << "CSR only support Edges of order \"srcdst\","
      << " but got \"" << order << "\".";
    const auto& coo = aten::CSRToCOO(adj_, false);
    return EdgeArray{coo.row, coo.col, coo.data};
  }

  uint64_t InDegree(dgl_type_t etype, dgl_id_t vid) const override {
642
    LOG(FATAL) << "Not enabled for CSR graph.";
643
644
645
646
    return {};
  }

  DegreeArray InDegrees(dgl_type_t etype, IdArray vids) const override {
647
    LOG(FATAL) << "Not enabled for CSR graph.";
648
649
650
651
    return {};
  }

  uint64_t OutDegree(dgl_type_t etype, dgl_id_t vid) const override {
Minjie Wang's avatar
Minjie Wang committed
652
    CHECK(HasVertex(SrcType(), vid)) << "Invalid src vertex id: " << vid;
653
654
655
656
    return aten::CSRGetRowNNZ(adj_, vid);
  }

  DegreeArray OutDegrees(dgl_type_t etype, IdArray vids) const override {
657
    CHECK(aten::IsValidIdArray(vids)) << "Invalid vertex id array.";
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
    return aten::CSRGetRowNNZ(adj_, vids);
  }

  DGLIdIters SuccVec(dgl_type_t etype, dgl_id_t vid) const override {
    // TODO(minjie): This still assumes the data type and device context
    //   of this graph. Should fix later.
    const dgl_id_t* indptr_data = static_cast<dgl_id_t*>(adj_.indptr->data);
    const dgl_id_t* indices_data = static_cast<dgl_id_t*>(adj_.indices->data);
    const dgl_id_t start = indptr_data[vid];
    const dgl_id_t end = indptr_data[vid + 1];
    return DGLIdIters(indices_data + start, indices_data + end);
  }

  DGLIdIters OutEdgeVec(dgl_type_t etype, dgl_id_t vid) const override {
    // TODO(minjie): This still assumes the data type and device context
    //   of this graph. Should fix later.
    const dgl_id_t* indptr_data = static_cast<dgl_id_t*>(adj_.indptr->data);
    const dgl_id_t* eid_data = static_cast<dgl_id_t*>(adj_.data->data);
    const dgl_id_t start = indptr_data[vid];
    const dgl_id_t end = indptr_data[vid + 1];
    return DGLIdIters(eid_data + start, eid_data + end);
  }

  DGLIdIters PredVec(dgl_type_t etype, dgl_id_t vid) const override {
682
    LOG(FATAL) << "Not enabled for CSR graph.";
683
684
685
686
    return {};
  }

  DGLIdIters InEdgeVec(dgl_type_t etype, dgl_id_t vid) const override {
687
    LOG(FATAL) << "Not enabled for CSR graph.";
688
689
690
691
692
693
694
695
696
    return {};
  }

  std::vector<IdArray> GetAdj(
      dgl_type_t etype, bool transpose, const std::string &fmt) const override {
    CHECK(!transpose && fmt == "csr") << "Not valid adj format request.";
    return {adj_.indptr, adj_.indices, adj_.data};
  }

697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
  aten::COOMatrix GetCOOMatrix(dgl_type_t etype) const override {
    LOG(FATAL) << "Not enabled for CSR graph";
    return aten::COOMatrix();
  }

  aten::CSRMatrix GetCSCMatrix(dgl_type_t etype) const override {
    LOG(FATAL) << "Not enabled for CSR graph";
    return aten::CSRMatrix();
  }

  aten::CSRMatrix GetCSRMatrix(dgl_type_t etype) const override {
    return adj_;
  }

  SparseFormat SelectFormat(dgl_type_t etype, SparseFormat preferred_format) const override {
    LOG(FATAL) << "Not enabled for CSR graph";
    return SparseFormat::ANY;
  }

716
  HeteroSubgraph VertexSubgraph(const std::vector<IdArray>& vids) const override {
Minjie Wang's avatar
Minjie Wang committed
717
718
719
720
    CHECK_EQ(vids.size(), NumVertexTypes()) << "Number of vertex types mismatch";
    auto srcvids = vids[SrcType()], dstvids = vids[DstType()];
    CHECK(aten::IsValidIdArray(srcvids)) << "Invalid vertex id array.";
    CHECK(aten::IsValidIdArray(dstvids)) << "Invalid vertex id array.";
721
    HeteroSubgraph subg;
Minjie Wang's avatar
Minjie Wang committed
722
    const auto& submat = aten::CSRSliceMatrix(adj_, srcvids, dstvids);
723
    IdArray sub_eids = aten::Range(0, submat.data->shape[0], NumBits(), Context());
Minjie Wang's avatar
Minjie Wang committed
724
    subg.graph = std::make_shared<CSR>(meta_graph(), submat.num_rows, submat.num_cols,
725
726
727
728
729
730
731
732
        submat.indptr, submat.indices, sub_eids);
    subg.induced_vertices = vids;
    subg.induced_edges.emplace_back(submat.data);
    return subg;
  }

  HeteroSubgraph EdgeSubgraph(
      const std::vector<IdArray>& eids, bool preserve_nodes = false) const override {
733
    LOG(FATAL) << "Not enabled for CSR graph.";
734
735
736
737
738
739
740
741
742
743
744
745
746
    return {};
  }

  aten::CSRMatrix adj() const {
    return adj_;
  }

 private:
  /*! \brief internal adjacency matrix. Data array stores edge ids */
  aten::CSRMatrix adj_;

  /*! \brief multi-graph flag */
  Lazy<bool> is_multigraph_;
Da Zheng's avatar
Da Zheng committed
747
748
749

  /*! \brief indicate that the edges are stored in the sorted order. */
  bool sorted_;
750
751
752
753
};

//////////////////////////////////////////////////////////
//
Minjie Wang's avatar
Minjie Wang committed
754
// unit graph implementation
755
756
757
//
//////////////////////////////////////////////////////////

758
759
760
761
DLDataType UnitGraph::DataType() const {
  return GetAny()->DataType();
}

Minjie Wang's avatar
Minjie Wang committed
762
DLContext UnitGraph::Context() const {
763
764
765
  return GetAny()->Context();
}

Minjie Wang's avatar
Minjie Wang committed
766
uint8_t UnitGraph::NumBits() const {
767
768
769
  return GetAny()->NumBits();
}

Minjie Wang's avatar
Minjie Wang committed
770
bool UnitGraph::IsMultigraph() const {
771
772
773
  return GetAny()->IsMultigraph();
}

Minjie Wang's avatar
Minjie Wang committed
774
uint64_t UnitGraph::NumVertices(dgl_type_t vtype) const {
775
776
777
778
779
  const SparseFormat fmt = SelectFormat(SparseFormat::ANY);
  const auto ptr = GetFormat(fmt);
  // TODO(BarclayII): we have a lot of special handling for CSC.
  // Need to have a UnitGraph::CSC backend instead.
  if (fmt == SparseFormat::CSC)
Minjie Wang's avatar
Minjie Wang committed
780
    vtype = (vtype == SrcType()) ? DstType() : SrcType();
781
  return ptr->NumVertices(vtype);
782
783
}

Minjie Wang's avatar
Minjie Wang committed
784
uint64_t UnitGraph::NumEdges(dgl_type_t etype) const {
785
786
787
  return GetAny()->NumEdges(etype);
}

Minjie Wang's avatar
Minjie Wang committed
788
bool UnitGraph::HasVertex(dgl_type_t vtype, dgl_id_t vid) const {
789
790
791
  const SparseFormat fmt = SelectFormat(SparseFormat::ANY);
  const auto ptr = GetFormat(fmt);
  if (fmt == SparseFormat::CSC)
Minjie Wang's avatar
Minjie Wang committed
792
    vtype = (vtype == SrcType()) ? DstType() : SrcType();
793
  return ptr->HasVertex(vtype, vid);
794
795
}

Minjie Wang's avatar
Minjie Wang committed
796
BoolArray UnitGraph::HasVertices(dgl_type_t vtype, IdArray vids) const {
797
  CHECK(aten::IsValidIdArray(vids)) << "Invalid id array input";
798
799
800
  return aten::LT(vids, NumVertices(vtype));
}

Minjie Wang's avatar
Minjie Wang committed
801
bool UnitGraph::HasEdgeBetween(dgl_type_t etype, dgl_id_t src, dgl_id_t dst) const {
802
803
804
805
806
807
  const SparseFormat fmt = SelectFormat(SparseFormat::ANY);
  const auto ptr = GetFormat(fmt);
  if (fmt == SparseFormat::CSC)
    return ptr->HasEdgeBetween(etype, dst, src);
  else
    return ptr->HasEdgeBetween(etype, src, dst);
808
809
}

Minjie Wang's avatar
Minjie Wang committed
810
BoolArray UnitGraph::HasEdgesBetween(
811
    dgl_type_t etype, IdArray src, IdArray dst) const {
812
813
814
815
816
817
  const SparseFormat fmt = SelectFormat(SparseFormat::ANY);
  const auto ptr = GetFormat(fmt);
  if (fmt == SparseFormat::CSC)
    return ptr->HasEdgesBetween(etype, dst, src);
  else
    return ptr->HasEdgesBetween(etype, src, dst);
818
819
}

Minjie Wang's avatar
Minjie Wang committed
820
IdArray UnitGraph::Predecessors(dgl_type_t etype, dgl_id_t dst) const {
821
822
823
824
825
826
  const SparseFormat fmt = SelectFormat(SparseFormat::CSC);
  const auto ptr = GetFormat(fmt);
  if (fmt == SparseFormat::CSC)
    return ptr->Successors(etype, dst);
  else
    return ptr->Predecessors(etype, dst);
827
828
}

Minjie Wang's avatar
Minjie Wang committed
829
IdArray UnitGraph::Successors(dgl_type_t etype, dgl_id_t src) const {
830
831
832
  const SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  const auto ptr = GetFormat(fmt);
  return ptr->Successors(etype, src);
833
834
}

Minjie Wang's avatar
Minjie Wang committed
835
IdArray UnitGraph::EdgeId(dgl_type_t etype, dgl_id_t src, dgl_id_t dst) const {
836
837
838
839
840
841
  const SparseFormat fmt = SelectFormat(SparseFormat::ANY);
  const auto ptr = GetFormat(fmt);
  if (fmt == SparseFormat::CSC)
    return ptr->EdgeId(etype, dst, src);
  else
    return ptr->EdgeId(etype, src, dst);
842
843
}

Minjie Wang's avatar
Minjie Wang committed
844
EdgeArray UnitGraph::EdgeIds(dgl_type_t etype, IdArray src, IdArray dst) const {
845
846
847
848
  const SparseFormat fmt = SelectFormat(SparseFormat::ANY);
  const auto ptr = GetFormat(fmt);
  if (fmt == SparseFormat::CSC) {
    EdgeArray edges = ptr->EdgeIds(etype, dst, src);
849
850
    return EdgeArray{edges.dst, edges.src, edges.id};
  } else {
851
    return ptr->EdgeIds(etype, src, dst);
852
853
854
  }
}

Minjie Wang's avatar
Minjie Wang committed
855
std::pair<dgl_id_t, dgl_id_t> UnitGraph::FindEdge(dgl_type_t etype, dgl_id_t eid) const {
856
857
858
  const SparseFormat fmt = SelectFormat(SparseFormat::COO);
  const auto ptr = GetFormat(fmt);
  return ptr->FindEdge(etype, eid);
859
860
}

Minjie Wang's avatar
Minjie Wang committed
861
EdgeArray UnitGraph::FindEdges(dgl_type_t etype, IdArray eids) const {
862
863
864
  const SparseFormat fmt = SelectFormat(SparseFormat::COO);
  const auto ptr = GetFormat(fmt);
  return ptr->FindEdges(etype, eids);
865
866
}

Minjie Wang's avatar
Minjie Wang committed
867
EdgeArray UnitGraph::InEdges(dgl_type_t etype, dgl_id_t vid) const {
868
869
870
871
872
873
874
875
  const SparseFormat fmt = SelectFormat(SparseFormat::CSC);
  const auto ptr = GetFormat(fmt);
  if (fmt == SparseFormat::CSC) {
    const EdgeArray& ret = ptr->OutEdges(etype, vid);
    return {ret.dst, ret.src, ret.id};
  } else {
    return ptr->InEdges(etype, vid);
  }
876
877
}

Minjie Wang's avatar
Minjie Wang committed
878
EdgeArray UnitGraph::InEdges(dgl_type_t etype, IdArray vids) const {
879
880
881
882
883
884
885
886
  const SparseFormat fmt = SelectFormat(SparseFormat::CSC);
  const auto ptr = GetFormat(fmt);
  if (fmt == SparseFormat::CSC) {
    const EdgeArray& ret = ptr->OutEdges(etype, vids);
    return {ret.dst, ret.src, ret.id};
  } else {
    return ptr->InEdges(etype, vids);
  }
887
888
}

Minjie Wang's avatar
Minjie Wang committed
889
EdgeArray UnitGraph::OutEdges(dgl_type_t etype, dgl_id_t vid) const {
890
891
892
  const SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  const auto ptr = GetFormat(fmt);
  return ptr->OutEdges(etype, vid);
893
894
}

Minjie Wang's avatar
Minjie Wang committed
895
EdgeArray UnitGraph::OutEdges(dgl_type_t etype, IdArray vids) const {
896
897
898
  const SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  const auto ptr = GetFormat(fmt);
  return ptr->OutEdges(etype, vids);
899
900
}

Minjie Wang's avatar
Minjie Wang committed
901
EdgeArray UnitGraph::Edges(dgl_type_t etype, const std::string &order) const {
902
903
904
905
  SparseFormat fmt;
  if (order == std::string("eid")) {
    fmt = SelectFormat(SparseFormat::COO);
  } else if (order.empty()) {
906
    // arbitrary order
907
    fmt = SelectFormat(SparseFormat::ANY);
908
  } else if (order == std::string("srcdst")) {
909
    fmt = SelectFormat(SparseFormat::CSR);
910
911
  } else {
    LOG(FATAL) << "Unsupported order request: " << order;
912
    return {};
913
  }
914
915
916
917
918
919

  const auto& edges = GetFormat(fmt)->Edges(etype, order);
  if (fmt == SparseFormat::CSC)
    return EdgeArray{edges.dst, edges.src, edges.id};
  else
    return edges;
920
921
}

Minjie Wang's avatar
Minjie Wang committed
922
uint64_t UnitGraph::InDegree(dgl_type_t etype, dgl_id_t vid) const {
923
924
925
926
927
928
  SparseFormat fmt = SelectFormat(SparseFormat::CSC);
  const auto ptr = GetFormat(fmt);
  if (fmt == SparseFormat::CSC)
    return ptr->OutDegree(etype, vid);
  else
    return ptr->InDegree(etype, vid);
929
930
}

Minjie Wang's avatar
Minjie Wang committed
931
DegreeArray UnitGraph::InDegrees(dgl_type_t etype, IdArray vids) const {
932
933
934
935
936
937
  SparseFormat fmt = SelectFormat(SparseFormat::CSC);
  const auto ptr = GetFormat(fmt);
  if (fmt == SparseFormat::CSC)
    return ptr->OutDegrees(etype, vids);
  else
    return ptr->InDegrees(etype, vids);
938
939
}

Minjie Wang's avatar
Minjie Wang committed
940
uint64_t UnitGraph::OutDegree(dgl_type_t etype, dgl_id_t vid) const {
941
942
943
  SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  const auto ptr = GetFormat(fmt);
  return ptr->OutDegree(etype, vid);
944
945
}

Minjie Wang's avatar
Minjie Wang committed
946
DegreeArray UnitGraph::OutDegrees(dgl_type_t etype, IdArray vids) const {
947
948
949
  SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  const auto ptr = GetFormat(fmt);
  return ptr->OutDegrees(etype, vids);
950
951
}

Minjie Wang's avatar
Minjie Wang committed
952
DGLIdIters UnitGraph::SuccVec(dgl_type_t etype, dgl_id_t vid) const {
953
954
955
  SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  const auto ptr = GetFormat(fmt);
  return ptr->SuccVec(etype, vid);
956
957
}

Minjie Wang's avatar
Minjie Wang committed
958
DGLIdIters UnitGraph::OutEdgeVec(dgl_type_t etype, dgl_id_t vid) const {
959
960
961
  SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  const auto ptr = GetFormat(fmt);
  return ptr->OutEdgeVec(etype, vid);
962
963
}

Minjie Wang's avatar
Minjie Wang committed
964
DGLIdIters UnitGraph::PredVec(dgl_type_t etype, dgl_id_t vid) const {
965
966
967
968
969
970
  SparseFormat fmt = SelectFormat(SparseFormat::CSC);
  const auto ptr = GetFormat(fmt);
  if (fmt == SparseFormat::CSC)
    return ptr->SuccVec(etype, vid);
  else
    return ptr->PredVec(etype, vid);
971
972
}

Minjie Wang's avatar
Minjie Wang committed
973
DGLIdIters UnitGraph::InEdgeVec(dgl_type_t etype, dgl_id_t vid) const {
974
975
976
977
978
979
  SparseFormat fmt = SelectFormat(SparseFormat::CSC);
  const auto ptr = GetFormat(fmt);
  if (fmt == SparseFormat::CSC)
    return ptr->OutEdgeVec(etype, vid);
  else
    return ptr->InEdgeVec(etype, vid);
980
981
}

Minjie Wang's avatar
Minjie Wang committed
982
std::vector<IdArray> UnitGraph::GetAdj(
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
    dgl_type_t etype, bool transpose, const std::string &fmt) const {
  // TODO(minjie): Our current semantics of adjacency matrix is row for dst nodes and col for
  //   src nodes. Therefore, we need to flip the transpose flag. For example, transpose=False
  //   is equal to in edge CSR.
  //   We have this behavior because previously we use framework's SPMM and we don't cache
  //   reverse adj. This is not intuitive and also not consistent with networkx's
  //   to_scipy_sparse_matrix. With the upcoming custom kernel change, we should change the
  //   behavior and make row for src and col for dst.
  if (fmt == std::string("csr")) {
    return transpose? GetOutCSR()->GetAdj(etype, false, "csr")
      : GetInCSR()->GetAdj(etype, false, "csr");
  } else if (fmt == std::string("coo")) {
    return GetCOO()->GetAdj(etype, !transpose, fmt);
  } else {
    LOG(FATAL) << "unsupported adjacency matrix format: " << fmt;
    return {};
  }
}

Minjie Wang's avatar
Minjie Wang committed
1002
HeteroSubgraph UnitGraph::VertexSubgraph(const std::vector<IdArray>& vids) const {
1003
  // We prefer to generate a subgraph from out-csr.
1004
1005
  SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  HeteroSubgraph sg = GetFormat(fmt)->VertexSubgraph(vids);
1006
1007
  CSRPtr subcsr = std::dynamic_pointer_cast<CSR>(sg.graph);
  HeteroSubgraph ret;
Minjie Wang's avatar
Minjie Wang committed
1008
  ret.graph = HeteroGraphPtr(new UnitGraph(meta_graph(), nullptr, subcsr, nullptr));
1009
1010
1011
1012
1013
  ret.induced_vertices = std::move(sg.induced_vertices);
  ret.induced_edges = std::move(sg.induced_edges);
  return ret;
}

Minjie Wang's avatar
Minjie Wang committed
1014
HeteroSubgraph UnitGraph::EdgeSubgraph(
1015
    const std::vector<IdArray>& eids, bool preserve_nodes) const {
1016
1017
  SparseFormat fmt = SelectFormat(SparseFormat::COO);
  auto sg = GetFormat(fmt)->EdgeSubgraph(eids, preserve_nodes);
1018
1019
  COOPtr subcoo = std::dynamic_pointer_cast<COO>(sg.graph);
  HeteroSubgraph ret;
Minjie Wang's avatar
Minjie Wang committed
1020
  ret.graph = HeteroGraphPtr(new UnitGraph(meta_graph(), nullptr, nullptr, subcoo));
1021
1022
1023
1024
1025
  ret.induced_vertices = std::move(sg.induced_vertices);
  ret.induced_edges = std::move(sg.induced_edges);
  return ret;
}

Minjie Wang's avatar
Minjie Wang committed
1026
HeteroGraphPtr UnitGraph::CreateFromCOO(
1027
1028
    int64_t num_vtypes, int64_t num_src, int64_t num_dst,
    IdArray row, IdArray col,
1029
    SparseFormat restrict_format) {
Minjie Wang's avatar
Minjie Wang committed
1030
1031
1032
1033
1034
  CHECK(num_vtypes == 1 || num_vtypes == 2);
  if (num_vtypes == 1)
    CHECK_EQ(num_src, num_dst);
  auto mg = CreateUnitGraphMetaGraph(num_vtypes);
  COOPtr coo(new COO(mg, num_src, num_dst, row, col));
1035
1036
1037

  return HeteroGraphPtr(
      new UnitGraph(mg, nullptr, nullptr, coo, restrict_format));
1038
1039
}

1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
HeteroGraphPtr UnitGraph::CreateFromCOO(
    int64_t num_vtypes, const aten::COOMatrix& mat,
    SparseFormat restrict_format) {
  CHECK(num_vtypes == 1 || num_vtypes == 2);
  if (num_vtypes == 1)
    CHECK_EQ(mat.num_rows, mat.num_cols);
  auto mg = CreateUnitGraphMetaGraph(num_vtypes);
  COOPtr coo(new COO(mg, mat));
  return HeteroGraphPtr(
      new UnitGraph(mg, nullptr, nullptr, coo, restrict_format));
}

Minjie Wang's avatar
Minjie Wang committed
1052
1053
HeteroGraphPtr UnitGraph::CreateFromCSR(
    int64_t num_vtypes, int64_t num_src, int64_t num_dst,
1054
    IdArray indptr, IdArray indices, IdArray edge_ids, SparseFormat restrict_format) {
Minjie Wang's avatar
Minjie Wang committed
1055
1056
1057
1058
1059
  CHECK(num_vtypes == 1 || num_vtypes == 2);
  if (num_vtypes == 1)
    CHECK_EQ(num_src, num_dst);
  auto mg = CreateUnitGraphMetaGraph(num_vtypes);
  CSRPtr csr(new CSR(mg, num_src, num_dst, indptr, indices, edge_ids));
1060
  return HeteroGraphPtr(new UnitGraph(mg, nullptr, csr, nullptr, restrict_format));
1061
1062
}

1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
HeteroGraphPtr UnitGraph::CreateFromCSR(
    int64_t num_vtypes, const aten::CSRMatrix& mat,
    SparseFormat restrict_format) {
  CHECK(num_vtypes == 1 || num_vtypes == 2);
  if (num_vtypes == 1)
    CHECK_EQ(mat.num_rows, mat.num_cols);
  auto mg = CreateUnitGraphMetaGraph(num_vtypes);
  CSRPtr csr(new CSR(mg, mat));
  return HeteroGraphPtr(new UnitGraph(mg, nullptr, csr, nullptr, restrict_format));
}

Minjie Wang's avatar
Minjie Wang committed
1074
HeteroGraphPtr UnitGraph::AsNumBits(HeteroGraphPtr g, uint8_t bits) {
1075
1076
1077
1078
1079
1080
1081
  if (g->NumBits() == bits) {
    return g;
  } else {
    // TODO(minjie): since we don't have int32 operations,
    //   we make sure that this graph (on CPU) has materialized CSR,
    //   and then copy them to other context (usually GPU). This should
    //   be fixed later.
Minjie Wang's avatar
Minjie Wang committed
1082
    auto bg = std::dynamic_pointer_cast<UnitGraph>(g);
1083
    CHECK_NOTNULL(bg);
1084

1085
1086
    CSRPtr new_incsr = CSRPtr(new CSR(bg->GetInCSR()->AsNumBits(bits)));
    CSRPtr new_outcsr = CSRPtr(new CSR(bg->GetOutCSR()->AsNumBits(bits)));
1087
1088
    return HeteroGraphPtr(
        new UnitGraph(g->meta_graph(), new_incsr, new_outcsr, nullptr, bg->restrict_format_));
1089
1090
1091
  }
}

Minjie Wang's avatar
Minjie Wang committed
1092
HeteroGraphPtr UnitGraph::CopyTo(HeteroGraphPtr g, const DLContext& ctx) {
1093
1094
1095
1096
1097
1098
1099
  if (ctx == g->Context()) {
    return g;
  }
  // TODO(minjie): since we don't have GPU implementation of COO<->CSR,
  //   we make sure that this graph (on CPU) has materialized CSR,
  //   and then copy them to other context (usually GPU). This should
  //   be fixed later.
Minjie Wang's avatar
Minjie Wang committed
1100
  auto bg = std::dynamic_pointer_cast<UnitGraph>(g);
1101
  CHECK_NOTNULL(bg);
1102

1103
1104
  CSRPtr new_incsr = CSRPtr(new CSR(bg->GetInCSR()->CopyTo(ctx)));
  CSRPtr new_outcsr = CSRPtr(new CSR(bg->GetOutCSR()->CopyTo(ctx)));
1105
1106
  return HeteroGraphPtr(
      new UnitGraph(g->meta_graph(), new_incsr, new_outcsr, nullptr, bg->restrict_format_));
1107
1108
}

1109
1110
UnitGraph::UnitGraph(GraphPtr metagraph, CSRPtr in_csr, CSRPtr out_csr, COOPtr coo,
                     SparseFormat restrict_format)
Minjie Wang's avatar
Minjie Wang committed
1111
  : BaseHeteroGraph(metagraph), in_csr_(in_csr), out_csr_(out_csr), coo_(coo) {
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
  restrict_format_ = restrict_format;

  // If the graph is hypersparse and in COO format, switch the restricted format to COO.
  // If the graph is given as CSR, the indptr array is already materialized so we don't
  // care about restricting conversion anyway (even if it is hypersparse).
  if (restrict_format == SparseFormat::ANY) {
    if (coo && coo->IsHypersparse())
      restrict_format_ = SparseFormat::COO;
  }

1122
1123
1124
  CHECK(GetAny()) << "At least one graph structure should exist.";
}

Minjie Wang's avatar
Minjie Wang committed
1125
UnitGraph::CSRPtr UnitGraph::GetInCSR() const {
1126
1127
1128
  if (!in_csr_) {
    if (out_csr_) {
      const auto& newadj = aten::CSRTranspose(out_csr_->adj());
Minjie Wang's avatar
Minjie Wang committed
1129
      const_cast<UnitGraph*>(this)->in_csr_ = std::make_shared<CSR>(meta_graph(), newadj);
1130
1131
1132
1133
1134
    } else {
      CHECK(coo_) << "None of CSR, COO exist";
      const auto& adj = coo_->adj();
      const auto& newadj = aten::COOToCSR(
          aten::COOMatrix{adj.num_cols, adj.num_rows, adj.col, adj.row});
Minjie Wang's avatar
Minjie Wang committed
1135
      const_cast<UnitGraph*>(this)->in_csr_ = std::make_shared<CSR>(meta_graph(), newadj);
1136
1137
1138
1139
1140
1141
    }
  }
  return in_csr_;
}

/* !\brief Return out csr. If not exist, transpose the other one.*/
Minjie Wang's avatar
Minjie Wang committed
1142
UnitGraph::CSRPtr UnitGraph::GetOutCSR() const {
1143
1144
1145
  if (!out_csr_) {
    if (in_csr_) {
      const auto& newadj = aten::CSRTranspose(in_csr_->adj());
Minjie Wang's avatar
Minjie Wang committed
1146
      const_cast<UnitGraph*>(this)->out_csr_ = std::make_shared<CSR>(meta_graph(), newadj);
1147
1148
1149
    } else {
      CHECK(coo_) << "None of CSR, COO exist";
      const auto& newadj = aten::COOToCSR(coo_->adj());
Minjie Wang's avatar
Minjie Wang committed
1150
      const_cast<UnitGraph*>(this)->out_csr_ = std::make_shared<CSR>(meta_graph(), newadj);
1151
1152
1153
1154
1155
1156
    }
  }
  return out_csr_;
}

/* !\brief Return coo. If not exist, create from csr.*/
Minjie Wang's avatar
Minjie Wang committed
1157
UnitGraph::COOPtr UnitGraph::GetCOO() const {
1158
1159
  if (!coo_) {
    if (in_csr_) {
1160
1161
      const auto& newadj = aten::COOTranspose(aten::CSRToCOO(in_csr_->adj(), true));
      const_cast<UnitGraph*>(this)->coo_ = std::make_shared<COO>(meta_graph(), newadj);
1162
1163
1164
    } else {
      CHECK(out_csr_) << "Both CSR are missing.";
      const auto& newadj = aten::CSRToCOO(out_csr_->adj(), true);
Minjie Wang's avatar
Minjie Wang committed
1165
      const_cast<UnitGraph*>(this)->coo_ = std::make_shared<COO>(meta_graph(), newadj);
1166
1167
1168
1169
1170
    }
  }
  return coo_;
}

1171
aten::CSRMatrix UnitGraph::GetCSCMatrix(dgl_type_t etype) const {
1172
1173
1174
  return GetInCSR()->adj();
}

1175
aten::CSRMatrix UnitGraph::GetCSRMatrix(dgl_type_t etype) const {
1176
1177
1178
  return GetOutCSR()->adj();
}

1179
aten::COOMatrix UnitGraph::GetCOOMatrix(dgl_type_t etype) const {
1180
1181
1182
  return GetCOO()->adj();
}

Minjie Wang's avatar
Minjie Wang committed
1183
HeteroGraphPtr UnitGraph::GetAny() const {
1184
1185
1186
1187
1188
1189
1190
1191
1192
  if (in_csr_) {
    return in_csr_;
  } else if (out_csr_) {
    return out_csr_;
  } else {
    return coo_;
  }
}

1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
HeteroGraphPtr UnitGraph::GetFormat(SparseFormat format) const {
  switch (format) {
  case SparseFormat::CSR:
    return GetOutCSR();
  case SparseFormat::CSC:
    return GetInCSR();
  case SparseFormat::COO:
    return GetCOO();
  case SparseFormat::ANY:
    return GetAny();
  default:
    LOG(FATAL) << "unsupported format code";
    return nullptr;
  }
}

SparseFormat UnitGraph::SelectFormat(SparseFormat preferred_format) const {
  if (restrict_format_ != SparseFormat::ANY)
    return restrict_format_;
  else if (preferred_format != SparseFormat::ANY)
    return preferred_format;
  else if (in_csr_)
    return SparseFormat::CSC;
  else if (out_csr_)
    return SparseFormat::CSR;
  else
    return SparseFormat::COO;
}

1222
1223
1224
1225
1226
1227
1228
1229
UnitGraph* UnitGraph::EmptyGraph() {
  auto src = NewIdArray(0);
  auto dst = NewIdArray(0);
  auto mg = CreateUnitGraphMetaGraph(1);
  COOPtr coo(new COO(mg, 0, 0, src, dst));
  return new UnitGraph(mg, nullptr, nullptr, coo);
}

1230
1231
constexpr uint64_t kDGLSerialize_UnitGraphMagic = 0xDD2E60F0F6B4A127;

1232
// Using OurCSR
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
bool UnitGraph::Load(dmlc::Stream* fs) {
  uint64_t magicNum;
  CHECK(fs->Read(&magicNum)) << "Invalid Magic Number";
  CHECK_EQ(magicNum, kDGLSerialize_UnitGraphMagic) << "Invalid UnitGraph Data";
  uint64_t num_vtypes, num_src, num_dst;
  CHECK(fs->Read(&num_vtypes)) << "Invalid num_vtypes";
  CHECK(fs->Read(&num_src)) << "Invalid num_src";
  CHECK(fs->Read(&num_dst)) << "Invalid num_dst";
  aten::CSRMatrix csr_matrix;
  CHECK(fs->Read(&csr_matrix)) << "Invalid csr_matrix";
  auto mg = CreateUnitGraphMetaGraph(num_vtypes);
1244
1245
1246
  CSRPtr csr(new CSR(mg, num_src, num_dst, csr_matrix.indptr,
                     csr_matrix.indices, csr_matrix.data));
  *this = UnitGraph(mg, nullptr, csr, nullptr);
1247
1248
1249
  return true;
}

1250
// Using Out CSR
1251
1252
void UnitGraph::Save(dmlc::Stream* fs) const {
  // Following CreateFromCSR signature
1253
  aten::CSRMatrix csr_matrix = GetCSRMatrix(0);
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
  uint64_t num_vtypes = NumVertexTypes();
  uint64_t num_src = NumVertices(SrcType());
  uint64_t num_dst = NumVertices(DstType());
  fs->Write(kDGLSerialize_UnitGraphMagic);
  fs->Write(num_vtypes);
  fs->Write(num_src);
  fs->Write(num_dst);
  fs->Write(csr_matrix);
}

1264
}  // namespace dgl