"src/vscode:/vscode.git/clone" did not exist on "dc277501c74673b71813563ab58ba2a28a58ba2f"
unit_graph.cc 42.6 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
 */
#include <dgl/array.h>
7
#include <dgl/base_heterograph.h>
8
9
#include <dgl/immutable_graph.h>
#include <dgl/lazy.h>
10
11

#include "../c_api_common.h"
12
#include "./unit_graph.h"
13
14

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
  bool Load(dmlc::Stream* fs) {
    auto meta_imgraph = Serializer::make_shared<ImmutableGraph>();
    CHECK(fs->Read(&meta_imgraph)) << "Invalid meta graph";
    meta_graph_ = meta_imgraph;
    CHECK(fs->Read(&adj_)) << "Invalid adj matrix";
    return true;
  }
  void Save(dmlc::Stream* fs) const {
    auto meta_graph_ptr = ImmutableGraph::ToImmutable(meta_graph());
    fs->Write(meta_graph_ptr);
    fs->Write(adj_);
  }

417
 private:
418
419
420
421
  friend class Serializer;

  COO() {}

422
423
424
425
426
427
428
429
430
431
432
433
434
435
  /*! \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
436
class UnitGraph::CSR : public BaseHeteroGraph {
437
 public:
Minjie Wang's avatar
Minjie Wang committed
438
  CSR(GraphPtr metagraph, int64_t num_src, int64_t num_dst,
439
      IdArray indptr, IdArray indices, IdArray edge_ids)
Minjie Wang's avatar
Minjie Wang committed
440
    : BaseHeteroGraph(metagraph) {
441
442
443
444
445
    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";
446
447
448
    adj_ = aten::CSRMatrix{num_src, num_dst, indptr, indices, edge_ids};
  }

Minjie Wang's avatar
Minjie Wang committed
449
  CSR(GraphPtr metagraph, int64_t num_src, int64_t num_dst,
450
      IdArray indptr, IdArray indices, IdArray edge_ids, bool is_multigraph)
Minjie Wang's avatar
Minjie Wang committed
451
    : BaseHeteroGraph(metagraph), is_multigraph_(is_multigraph) {
452
453
454
455
456
    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";
457
458
459
    adj_ = aten::CSRMatrix{num_src, num_dst, indptr, indices, edge_ids};
  }

460
  CSR(GraphPtr metagraph, const aten::CSRMatrix& csr)
Da Zheng's avatar
Da Zheng committed
461
462
    : BaseHeteroGraph(metagraph), adj_(csr) {
  }
463

Minjie Wang's avatar
Minjie Wang committed
464
465
  inline dgl_type_t SrcType() const {
    return 0;
466
  }
Minjie Wang's avatar
Minjie Wang committed
467
468
469
470
471
472
473

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

  inline dgl_type_t EdgeType() const {
    return 0;
474
475
476
  }

  HeteroGraphPtr GetRelationGraph(dgl_type_t etype) const override {
Minjie Wang's avatar
Minjie Wang committed
477
    LOG(FATAL) << "The method shouldn't be called for UnitGraph graph. "
478
479
480
481
482
      << "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
483
    LOG(FATAL) << "UnitGraph graph is not mutable.";
484
485
486
  }

  void AddEdge(dgl_type_t etype, dgl_id_t src, dgl_id_t dst) override {
Minjie Wang's avatar
Minjie Wang committed
487
    LOG(FATAL) << "UnitGraph graph is not mutable.";
488
489
490
  }

  void AddEdges(dgl_type_t etype, IdArray src_ids, IdArray dst_ids) override {
Minjie Wang's avatar
Minjie Wang committed
491
    LOG(FATAL) << "UnitGraph graph is not mutable.";
492
493
494
  }

  void Clear() override {
Minjie Wang's avatar
Minjie Wang committed
495
    LOG(FATAL) << "UnitGraph graph is not mutable.";
496
497
  }

498
499
500
501
  DLDataType DataType() const override {
    return adj_.indices->dtype;
  }

502
503
504
505
506
507
508
509
  DLContext Context() const override {
    return adj_.indices->ctx;
  }

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

510
511
512
513
514
  CSR AsNumBits(uint8_t bits) const {
    if (NumBits() == bits) {
      return *this;
    } else {
      CSR ret(
Minjie Wang's avatar
Minjie Wang committed
515
          meta_graph_,
516
517
518
519
520
521
522
523
524
525
526
527
528
529
          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
530
          meta_graph_,
531
532
533
534
535
536
537
538
539
          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;
    }
  }

540
541
542
543
544
545
546
547
548
549
550
  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
551
    if (vtype == SrcType()) {
552
      return adj_.num_rows;
Minjie Wang's avatar
Minjie Wang committed
553
    } else if (vtype == DstType()) {
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
      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
575
576
    CHECK(HasVertex(SrcType(), src)) << "Invalid src vertex id: " << src;
    CHECK(HasVertex(DstType(), dst)) << "Invalid dst vertex id: " << dst;
577
578
579
580
    return aten::CSRIsNonZero(adj_, src, dst);
  }

  BoolArray HasEdgesBetween(dgl_type_t etype, IdArray src_ids, IdArray dst_ids) const override {
581
582
    CHECK(aten::IsValidIdArray(src_ids)) << "Invalid vertex id array.";
    CHECK(aten::IsValidIdArray(dst_ids)) << "Invalid vertex id array.";
583
584
585
586
587
588
589
590
591
    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
592
    CHECK(HasVertex(SrcType(), src)) << "Invalid src vertex id: " << src;
593
594
595
596
    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
597
598
    CHECK(HasVertex(SrcType(), src)) << "Invalid src vertex id: " << src;
    CHECK(HasVertex(DstType(), dst)) << "Invalid dst vertex id: " << dst;
599
600
601
602
    return aten::CSRGetData(adj_, src, dst);
  }

  EdgeArray EdgeIds(dgl_type_t etype, IdArray src, IdArray dst) const override {
603
604
    CHECK(aten::IsValidIdArray(src)) << "Invalid vertex id array.";
    CHECK(aten::IsValidIdArray(dst)) << "Invalid vertex id array.";
605
606
607
608
609
    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 {
610
    LOG(FATAL) << "Not enabled for CSR graph.";
611
612
613
614
    return {};
  }

  EdgeArray FindEdges(dgl_type_t etype, IdArray eids) const override {
615
    LOG(FATAL) << "Not enabled for CSR graph.";
616
617
618
619
    return {};
  }

  EdgeArray InEdges(dgl_type_t etype, dgl_id_t vid) const override {
620
    LOG(FATAL) << "Not enabled for CSR graph.";
621
622
623
624
    return {};
  }

  EdgeArray InEdges(dgl_type_t etype, IdArray vids) const override {
625
    LOG(FATAL) << "Not enabled for CSR graph.";
626
627
628
629
    return {};
  }

  EdgeArray OutEdges(dgl_type_t etype, dgl_id_t vid) const override {
Minjie Wang's avatar
Minjie Wang committed
630
    CHECK(HasVertex(SrcType(), vid)) << "Invalid src vertex id: " << vid;
631
632
633
634
635
636
637
    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 {
638
    CHECK(aten::IsValidIdArray(vids)) << "Invalid vertex id array.";
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
    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 {
656
    LOG(FATAL) << "Not enabled for CSR graph.";
657
658
659
660
    return {};
  }

  DegreeArray InDegrees(dgl_type_t etype, IdArray vids) const override {
661
    LOG(FATAL) << "Not enabled for CSR graph.";
662
663
664
665
    return {};
  }

  uint64_t OutDegree(dgl_type_t etype, dgl_id_t vid) const override {
Minjie Wang's avatar
Minjie Wang committed
666
    CHECK(HasVertex(SrcType(), vid)) << "Invalid src vertex id: " << vid;
667
668
669
670
    return aten::CSRGetRowNNZ(adj_, vid);
  }

  DegreeArray OutDegrees(dgl_type_t etype, IdArray vids) const override {
671
    CHECK(aten::IsValidIdArray(vids)) << "Invalid vertex id array.";
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
    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 {
696
    LOG(FATAL) << "Not enabled for CSR graph.";
697
698
699
700
    return {};
  }

  DGLIdIters InEdgeVec(dgl_type_t etype, dgl_id_t vid) const override {
701
    LOG(FATAL) << "Not enabled for CSR graph.";
702
703
704
705
706
707
708
709
710
    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};
  }

711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
  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;
  }

730
  HeteroSubgraph VertexSubgraph(const std::vector<IdArray>& vids) const override {
Minjie Wang's avatar
Minjie Wang committed
731
732
733
734
    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.";
735
    HeteroSubgraph subg;
Minjie Wang's avatar
Minjie Wang committed
736
    const auto& submat = aten::CSRSliceMatrix(adj_, srcvids, dstvids);
737
    IdArray sub_eids = aten::Range(0, submat.data->shape[0], NumBits(), Context());
Minjie Wang's avatar
Minjie Wang committed
738
    subg.graph = std::make_shared<CSR>(meta_graph(), submat.num_rows, submat.num_cols,
739
740
741
742
743
744
745
746
        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 {
747
    LOG(FATAL) << "Not enabled for CSR graph.";
748
749
750
751
752
753
754
    return {};
  }

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

755
756
757
758
759
760
761
762
763
764
765
766
767
768
  bool Load(dmlc::Stream* fs) {
    auto meta_imgraph = Serializer::make_shared<ImmutableGraph>();
    CHECK(fs->Read(&meta_imgraph)) << "Invalid meta graph";
    meta_graph_ = meta_imgraph;
    CHECK(fs->Read(&adj_)) << "Invalid adj matrix";
    return true;
  }
  void Save(dmlc::Stream* fs) const {
    auto meta_graph_ptr = ImmutableGraph::ToImmutable(meta_graph());
    fs->Write(meta_graph_ptr);
    fs->Write(adj_);
  }


769
 private:
770
771
772
773
  friend class Serializer;

  CSR() {};

774
775
776
777
778
779
780
781
782
  /*! \brief internal adjacency matrix. Data array stores edge ids */
  aten::CSRMatrix adj_;

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

//////////////////////////////////////////////////////////
//
Minjie Wang's avatar
Minjie Wang committed
783
// unit graph implementation
784
785
786
//
//////////////////////////////////////////////////////////

787
788
789
790
DLDataType UnitGraph::DataType() const {
  return GetAny()->DataType();
}

Minjie Wang's avatar
Minjie Wang committed
791
DLContext UnitGraph::Context() const {
792
793
794
  return GetAny()->Context();
}

Minjie Wang's avatar
Minjie Wang committed
795
uint8_t UnitGraph::NumBits() const {
796
797
798
  return GetAny()->NumBits();
}

Minjie Wang's avatar
Minjie Wang committed
799
bool UnitGraph::IsMultigraph() const {
800
801
802
  return GetAny()->IsMultigraph();
}

Minjie Wang's avatar
Minjie Wang committed
803
uint64_t UnitGraph::NumVertices(dgl_type_t vtype) const {
804
805
806
807
808
  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
809
    vtype = (vtype == SrcType()) ? DstType() : SrcType();
810
  return ptr->NumVertices(vtype);
811
812
}

Minjie Wang's avatar
Minjie Wang committed
813
uint64_t UnitGraph::NumEdges(dgl_type_t etype) const {
814
815
816
  return GetAny()->NumEdges(etype);
}

Minjie Wang's avatar
Minjie Wang committed
817
bool UnitGraph::HasVertex(dgl_type_t vtype, dgl_id_t vid) const {
818
819
820
  const SparseFormat fmt = SelectFormat(SparseFormat::ANY);
  const auto ptr = GetFormat(fmt);
  if (fmt == SparseFormat::CSC)
Minjie Wang's avatar
Minjie Wang committed
821
    vtype = (vtype == SrcType()) ? DstType() : SrcType();
822
  return ptr->HasVertex(vtype, vid);
823
824
}

Minjie Wang's avatar
Minjie Wang committed
825
BoolArray UnitGraph::HasVertices(dgl_type_t vtype, IdArray vids) const {
826
  CHECK(aten::IsValidIdArray(vids)) << "Invalid id array input";
827
828
829
  return aten::LT(vids, NumVertices(vtype));
}

Minjie Wang's avatar
Minjie Wang committed
830
bool UnitGraph::HasEdgeBetween(dgl_type_t etype, dgl_id_t src, dgl_id_t dst) const {
831
832
833
834
835
836
  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);
837
838
}

Minjie Wang's avatar
Minjie Wang committed
839
BoolArray UnitGraph::HasEdgesBetween(
840
    dgl_type_t etype, IdArray src, IdArray dst) const {
841
842
843
844
845
846
  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);
847
848
}

Minjie Wang's avatar
Minjie Wang committed
849
IdArray UnitGraph::Predecessors(dgl_type_t etype, dgl_id_t dst) const {
850
851
852
853
854
855
  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);
856
857
}

Minjie Wang's avatar
Minjie Wang committed
858
IdArray UnitGraph::Successors(dgl_type_t etype, dgl_id_t src) const {
859
860
861
  const SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  const auto ptr = GetFormat(fmt);
  return ptr->Successors(etype, src);
862
863
}

Minjie Wang's avatar
Minjie Wang committed
864
IdArray UnitGraph::EdgeId(dgl_type_t etype, dgl_id_t src, dgl_id_t dst) const {
865
866
867
868
869
870
  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);
871
872
}

Minjie Wang's avatar
Minjie Wang committed
873
EdgeArray UnitGraph::EdgeIds(dgl_type_t etype, IdArray src, IdArray dst) const {
874
875
876
877
  const SparseFormat fmt = SelectFormat(SparseFormat::ANY);
  const auto ptr = GetFormat(fmt);
  if (fmt == SparseFormat::CSC) {
    EdgeArray edges = ptr->EdgeIds(etype, dst, src);
878
879
    return EdgeArray{edges.dst, edges.src, edges.id};
  } else {
880
    return ptr->EdgeIds(etype, src, dst);
881
882
883
  }
}

Minjie Wang's avatar
Minjie Wang committed
884
std::pair<dgl_id_t, dgl_id_t> UnitGraph::FindEdge(dgl_type_t etype, dgl_id_t eid) const {
885
886
887
  const SparseFormat fmt = SelectFormat(SparseFormat::COO);
  const auto ptr = GetFormat(fmt);
  return ptr->FindEdge(etype, eid);
888
889
}

Minjie Wang's avatar
Minjie Wang committed
890
EdgeArray UnitGraph::FindEdges(dgl_type_t etype, IdArray eids) const {
891
892
893
  const SparseFormat fmt = SelectFormat(SparseFormat::COO);
  const auto ptr = GetFormat(fmt);
  return ptr->FindEdges(etype, eids);
894
895
}

Minjie Wang's avatar
Minjie Wang committed
896
EdgeArray UnitGraph::InEdges(dgl_type_t etype, dgl_id_t vid) const {
897
898
899
900
901
902
903
904
  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);
  }
905
906
}

Minjie Wang's avatar
Minjie Wang committed
907
EdgeArray UnitGraph::InEdges(dgl_type_t etype, IdArray vids) const {
908
909
910
911
912
913
914
915
  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);
  }
916
917
}

Minjie Wang's avatar
Minjie Wang committed
918
EdgeArray UnitGraph::OutEdges(dgl_type_t etype, dgl_id_t vid) const {
919
920
921
  const SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  const auto ptr = GetFormat(fmt);
  return ptr->OutEdges(etype, vid);
922
923
}

Minjie Wang's avatar
Minjie Wang committed
924
EdgeArray UnitGraph::OutEdges(dgl_type_t etype, IdArray vids) const {
925
926
927
  const SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  const auto ptr = GetFormat(fmt);
  return ptr->OutEdges(etype, vids);
928
929
}

Minjie Wang's avatar
Minjie Wang committed
930
EdgeArray UnitGraph::Edges(dgl_type_t etype, const std::string &order) const {
931
932
933
934
  SparseFormat fmt;
  if (order == std::string("eid")) {
    fmt = SelectFormat(SparseFormat::COO);
  } else if (order.empty()) {
935
    // arbitrary order
936
    fmt = SelectFormat(SparseFormat::ANY);
937
  } else if (order == std::string("srcdst")) {
938
    fmt = SelectFormat(SparseFormat::CSR);
939
940
  } else {
    LOG(FATAL) << "Unsupported order request: " << order;
941
    return {};
942
  }
943
944
945
946
947
948

  const auto& edges = GetFormat(fmt)->Edges(etype, order);
  if (fmt == SparseFormat::CSC)
    return EdgeArray{edges.dst, edges.src, edges.id};
  else
    return edges;
949
950
}

Minjie Wang's avatar
Minjie Wang committed
951
uint64_t UnitGraph::InDegree(dgl_type_t etype, dgl_id_t vid) const {
952
953
954
955
956
957
  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);
958
959
}

Minjie Wang's avatar
Minjie Wang committed
960
DegreeArray UnitGraph::InDegrees(dgl_type_t etype, IdArray vids) const {
961
962
963
964
965
966
  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);
967
968
}

Minjie Wang's avatar
Minjie Wang committed
969
uint64_t UnitGraph::OutDegree(dgl_type_t etype, dgl_id_t vid) const {
970
971
972
  SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  const auto ptr = GetFormat(fmt);
  return ptr->OutDegree(etype, vid);
973
974
}

Minjie Wang's avatar
Minjie Wang committed
975
DegreeArray UnitGraph::OutDegrees(dgl_type_t etype, IdArray vids) const {
976
977
978
  SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  const auto ptr = GetFormat(fmt);
  return ptr->OutDegrees(etype, vids);
979
980
}

Minjie Wang's avatar
Minjie Wang committed
981
DGLIdIters UnitGraph::SuccVec(dgl_type_t etype, dgl_id_t vid) const {
982
983
984
  SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  const auto ptr = GetFormat(fmt);
  return ptr->SuccVec(etype, vid);
985
986
}

Minjie Wang's avatar
Minjie Wang committed
987
DGLIdIters UnitGraph::OutEdgeVec(dgl_type_t etype, dgl_id_t vid) const {
988
989
990
  SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  const auto ptr = GetFormat(fmt);
  return ptr->OutEdgeVec(etype, vid);
991
992
}

Minjie Wang's avatar
Minjie Wang committed
993
DGLIdIters UnitGraph::PredVec(dgl_type_t etype, dgl_id_t vid) const {
994
995
996
997
998
999
  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);
1000
1001
}

Minjie Wang's avatar
Minjie Wang committed
1002
DGLIdIters UnitGraph::InEdgeVec(dgl_type_t etype, dgl_id_t vid) const {
1003
1004
1005
1006
1007
1008
  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);
1009
1010
}

Minjie Wang's avatar
Minjie Wang committed
1011
std::vector<IdArray> UnitGraph::GetAdj(
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
    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
1031
HeteroSubgraph UnitGraph::VertexSubgraph(const std::vector<IdArray>& vids) const {
1032
  // We prefer to generate a subgraph from out-csr.
1033
1034
  SparseFormat fmt = SelectFormat(SparseFormat::CSR);
  HeteroSubgraph sg = GetFormat(fmt)->VertexSubgraph(vids);
1035
1036
  CSRPtr subcsr = std::dynamic_pointer_cast<CSR>(sg.graph);
  HeteroSubgraph ret;
Minjie Wang's avatar
Minjie Wang committed
1037
  ret.graph = HeteroGraphPtr(new UnitGraph(meta_graph(), nullptr, subcsr, nullptr));
1038
1039
1040
1041
1042
  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
1043
HeteroSubgraph UnitGraph::EdgeSubgraph(
1044
    const std::vector<IdArray>& eids, bool preserve_nodes) const {
1045
1046
  SparseFormat fmt = SelectFormat(SparseFormat::COO);
  auto sg = GetFormat(fmt)->EdgeSubgraph(eids, preserve_nodes);
1047
1048
  COOPtr subcoo = std::dynamic_pointer_cast<COO>(sg.graph);
  HeteroSubgraph ret;
Minjie Wang's avatar
Minjie Wang committed
1049
  ret.graph = HeteroGraphPtr(new UnitGraph(meta_graph(), nullptr, nullptr, subcoo));
1050
1051
1052
1053
1054
  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
1055
HeteroGraphPtr UnitGraph::CreateFromCOO(
1056
1057
    int64_t num_vtypes, int64_t num_src, int64_t num_dst,
    IdArray row, IdArray col,
1058
    SparseFormat restrict_format) {
Minjie Wang's avatar
Minjie Wang committed
1059
1060
1061
1062
1063
  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));
1064
1065
1066

  return HeteroGraphPtr(
      new UnitGraph(mg, nullptr, nullptr, coo, restrict_format));
1067
1068
}

1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
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
1081
1082
HeteroGraphPtr UnitGraph::CreateFromCSR(
    int64_t num_vtypes, int64_t num_src, int64_t num_dst,
1083
    IdArray indptr, IdArray indices, IdArray edge_ids, SparseFormat restrict_format) {
Minjie Wang's avatar
Minjie Wang committed
1084
1085
1086
1087
1088
  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));
1089
  return HeteroGraphPtr(new UnitGraph(mg, nullptr, csr, nullptr, restrict_format));
1090
1091
}

1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
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
1103
HeteroGraphPtr UnitGraph::AsNumBits(HeteroGraphPtr g, uint8_t bits) {
1104
1105
1106
1107
1108
1109
1110
  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
1111
    auto bg = std::dynamic_pointer_cast<UnitGraph>(g);
1112
    CHECK_NOTNULL(bg);
1113

1114
1115
    CSRPtr new_incsr = CSRPtr(new CSR(bg->GetInCSR()->AsNumBits(bits)));
    CSRPtr new_outcsr = CSRPtr(new CSR(bg->GetOutCSR()->AsNumBits(bits)));
1116
1117
    return HeteroGraphPtr(
        new UnitGraph(g->meta_graph(), new_incsr, new_outcsr, nullptr, bg->restrict_format_));
1118
1119
1120
  }
}

Minjie Wang's avatar
Minjie Wang committed
1121
HeteroGraphPtr UnitGraph::CopyTo(HeteroGraphPtr g, const DLContext& ctx) {
1122
1123
1124
1125
1126
1127
1128
  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
1129
  auto bg = std::dynamic_pointer_cast<UnitGraph>(g);
1130
  CHECK_NOTNULL(bg);
1131

1132
1133
  CSRPtr new_incsr = CSRPtr(new CSR(bg->GetInCSR()->CopyTo(ctx)));
  CSRPtr new_outcsr = CSRPtr(new CSR(bg->GetOutCSR()->CopyTo(ctx)));
1134
1135
  return HeteroGraphPtr(
      new UnitGraph(g->meta_graph(), new_incsr, new_outcsr, nullptr, bg->restrict_format_));
1136
1137
}

1138
1139
UnitGraph::UnitGraph(GraphPtr metagraph, CSRPtr in_csr, CSRPtr out_csr, COOPtr coo,
                     SparseFormat restrict_format)
Minjie Wang's avatar
Minjie Wang committed
1140
  : BaseHeteroGraph(metagraph), in_csr_(in_csr), out_csr_(out_csr), coo_(coo) {
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
  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;
  }

1151
1152
1153
  CHECK(GetAny()) << "At least one graph structure should exist.";
}

Minjie Wang's avatar
Minjie Wang committed
1154
UnitGraph::CSRPtr UnitGraph::GetInCSR() const {
1155
1156
1157
  if (!in_csr_) {
    if (out_csr_) {
      const auto& newadj = aten::CSRTranspose(out_csr_->adj());
Minjie Wang's avatar
Minjie Wang committed
1158
      const_cast<UnitGraph*>(this)->in_csr_ = std::make_shared<CSR>(meta_graph(), newadj);
1159
1160
1161
1162
1163
    } 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
1164
      const_cast<UnitGraph*>(this)->in_csr_ = std::make_shared<CSR>(meta_graph(), newadj);
1165
1166
1167
1168
1169
1170
    }
  }
  return in_csr_;
}

/* !\brief Return out csr. If not exist, transpose the other one.*/
Minjie Wang's avatar
Minjie Wang committed
1171
UnitGraph::CSRPtr UnitGraph::GetOutCSR() const {
1172
1173
1174
  if (!out_csr_) {
    if (in_csr_) {
      const auto& newadj = aten::CSRTranspose(in_csr_->adj());
Minjie Wang's avatar
Minjie Wang committed
1175
      const_cast<UnitGraph*>(this)->out_csr_ = std::make_shared<CSR>(meta_graph(), newadj);
1176
1177
1178
    } else {
      CHECK(coo_) << "None of CSR, COO exist";
      const auto& newadj = aten::COOToCSR(coo_->adj());
Minjie Wang's avatar
Minjie Wang committed
1179
      const_cast<UnitGraph*>(this)->out_csr_ = std::make_shared<CSR>(meta_graph(), newadj);
1180
1181
1182
1183
1184
1185
    }
  }
  return out_csr_;
}

/* !\brief Return coo. If not exist, create from csr.*/
Minjie Wang's avatar
Minjie Wang committed
1186
UnitGraph::COOPtr UnitGraph::GetCOO() const {
1187
1188
  if (!coo_) {
    if (in_csr_) {
1189
1190
      const auto& newadj = aten::COOTranspose(aten::CSRToCOO(in_csr_->adj(), true));
      const_cast<UnitGraph*>(this)->coo_ = std::make_shared<COO>(meta_graph(), newadj);
1191
1192
1193
    } else {
      CHECK(out_csr_) << "Both CSR are missing.";
      const auto& newadj = aten::CSRToCOO(out_csr_->adj(), true);
Minjie Wang's avatar
Minjie Wang committed
1194
      const_cast<UnitGraph*>(this)->coo_ = std::make_shared<COO>(meta_graph(), newadj);
1195
1196
1197
1198
1199
    }
  }
  return coo_;
}

1200
aten::CSRMatrix UnitGraph::GetCSCMatrix(dgl_type_t etype) const {
1201
1202
1203
  return GetInCSR()->adj();
}

1204
aten::CSRMatrix UnitGraph::GetCSRMatrix(dgl_type_t etype) const {
1205
1206
1207
  return GetOutCSR()->adj();
}

1208
aten::COOMatrix UnitGraph::GetCOOMatrix(dgl_type_t etype) const {
1209
1210
1211
  return GetCOO()->adj();
}

Minjie Wang's avatar
Minjie Wang committed
1212
HeteroGraphPtr UnitGraph::GetAny() const {
1213
1214
1215
1216
1217
1218
1219
1220
1221
  if (in_csr_) {
    return in_csr_;
  } else if (out_csr_) {
    return out_csr_;
  } else {
    return coo_;
  }
}

1222
1223
HeteroGraphPtr UnitGraph::GetFormat(SparseFormat format) const {
  switch (format) {
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
    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;
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
  }
}

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;
}

1251
1252
1253
1254
1255
1256
constexpr uint64_t kDGLSerialize_UnitGraphMagic = 0xDD2E60F0F6B4A127;

bool UnitGraph::Load(dmlc::Stream* fs) {
  uint64_t magicNum;
  CHECK(fs->Read(&magicNum)) << "Invalid Magic Number";
  CHECK_EQ(magicNum, kDGLSerialize_UnitGraphMagic) << "Invalid UnitGraph Data";
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278

  int64_t format_code;
  CHECK(fs->Read(&format_code)) << "Invalid format";
  restrict_format_ = static_cast<SparseFormat>(format_code);

  switch (restrict_format_) {
    case SparseFormat::COO:
      fs->Read(&coo_);
      break;
    case SparseFormat::CSR:
      fs->Read(&out_csr_);
      break;
    case SparseFormat::CSC:
      fs->Read(&in_csr_);
      break;
    default:
      LOG(FATAL) << "unsupported format code";
      break;
  }

  meta_graph_ = GetAny()->meta_graph();

1279
1280
1281
  return true;
}

1282

1283
1284
void UnitGraph::Save(dmlc::Stream* fs) const {
  fs->Write(kDGLSerialize_UnitGraphMagic);
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
  // Didn't write UnitGraph::meta_graph_, since it's included in the underlying
  // sparse matrix
  auto avail_fmt = SelectFormat(SparseFormat::ANY);
  fs->Write(static_cast<int64_t>(avail_fmt));
  switch (avail_fmt) {
    case SparseFormat::COO:
      fs->Write(GetCOO());
      break;
    case SparseFormat::CSR:
      fs->Write(GetOutCSR());
      break;
    case SparseFormat::CSC:
      fs->Write(GetInCSR());
      break;
    default:
      LOG(FATAL) << "unsupported format code";
      break;
  }
1303
1304
}

1305
}  // namespace dgl