1 | |
---|
2 | #include "sparse.h" |
---|
3 | #include "PrismSparseGlob.h" |
---|
4 | #include <new> |
---|
5 | |
---|
6 | //------------------------------------------------------------------------------ |
---|
7 | |
---|
8 | // local function prototypes |
---|
9 | |
---|
10 | static void split_mdp_rec(DdManager *ddman, DdNode *dd, DdNode **ndvars, int num_ndvars, int level, DdNode **matrices); |
---|
11 | static void split_mdp_and_sub_mdp_rec(DdManager *ddman, DdNode *dd, DdNode *subdd, DdNode **ndvars, int num_ndvars, int level, DdNode **matrices, DdNode **submatrices); |
---|
12 | static void traverse_mtbdd_matr_rec(DdManager *ddman, DdNode *dd, DdNode **rvars, DdNode **cvars, int num_vars, int level, ODDNode *row, ODDNode *col, int r, int c, int code, bool transpose); |
---|
13 | static void traverse_mtbdd_vect_rec(DdManager *ddman, DdNode *dd, DdNode **vars, int num_vars, int level, ODDNode *odd, int i, int code); |
---|
14 | |
---|
15 | // global variables (used by local functions) |
---|
16 | static int count; |
---|
17 | static int *starts, *starts2; |
---|
18 | static int *actions; |
---|
19 | static RMSparseMatrix *rmsm; |
---|
20 | static CMSparseMatrix *cmsm; |
---|
21 | static RCSparseMatrix *rcsm; |
---|
22 | static CMSRSparseMatrix *cmsrsm; |
---|
23 | static CMSCSparseMatrix *cmscsm; |
---|
24 | static NDSparseMatrix *ndsm; |
---|
25 | |
---|
26 | |
---|
27 | //----------------------------------------------------------------------------------- |
---|
28 | |
---|
29 | // build nondeterministic (mdp) sparse matrix |
---|
30 | // throws std::bad_alloc on out-of-memory |
---|
31 | |
---|
32 | NDSparseMatrix *build_nd_sparse_matrix(DdManager *ddman, DdNode *mdp, DdNode **rvars, DdNode **cvars, int num_vars, DdNode **ndvars, int num_ndvars, ODDNode *odd) |
---|
33 | { |
---|
34 | int i, n, nm, nc, nnz, max, max2; |
---|
35 | DdNode *tmp = NULL, **matrices = NULL, **matrices_bdds = NULL; |
---|
36 | |
---|
37 | // try/catch for memory allocation/deallocation |
---|
38 | try { |
---|
39 | |
---|
40 | // create new data structure |
---|
41 | ndsm = NULL; ndsm = new NDSparseMatrix(); |
---|
42 | |
---|
43 | // get number of states from odd |
---|
44 | n = ndsm->n = odd->eoff+odd->toff; |
---|
45 | // get num of choices (prob. distributions) |
---|
46 | Cudd_Ref(mdp); |
---|
47 | tmp = DD_ThereExists(ddman, DD_Not(ddman, DD_Equals(ddman, mdp, 0)), cvars, num_vars); |
---|
48 | nc = ndsm->nc = (int)DD_GetNumMinterms(ddman, tmp, num_vars+num_ndvars); |
---|
49 | // get num of transitions |
---|
50 | nnz = ndsm->nnz = (int)DD_GetNumMinterms(ddman, mdp, num_vars*2+num_ndvars); |
---|
51 | // break the mdp mtbdd into several (nm) mtbdds |
---|
52 | tmp = DD_ThereExists(ddman, tmp, rvars, num_vars); |
---|
53 | nm = (int)DD_GetNumMinterms(ddman, tmp, num_ndvars); |
---|
54 | Cudd_RecursiveDeref(ddman, tmp); |
---|
55 | matrices = new DdNode*[nm]; |
---|
56 | count = 0; |
---|
57 | split_mdp_rec(ddman, mdp, ndvars, num_ndvars, 0, matrices); |
---|
58 | // and for each one create a bdd storing which rows/choices are non-empty |
---|
59 | matrices_bdds = new DdNode*[nm]; |
---|
60 | for (i = 0; i < nm; i++) { |
---|
61 | Cudd_Ref(matrices[i]); |
---|
62 | matrices_bdds[i] = DD_ThereExists(ddman, DD_Not(ddman, DD_Equals(ddman, matrices[i], 0)), cvars, num_vars); |
---|
63 | } |
---|
64 | |
---|
65 | // create arrays |
---|
66 | ndsm->non_zeros = new double[nnz]; |
---|
67 | ndsm->cols = new unsigned int[nnz]; |
---|
68 | starts = NULL; starts = new int[n+1]; |
---|
69 | starts2 = NULL; starts2 = new int[nc+1]; |
---|
70 | |
---|
71 | // first traverse mtbdds to compute how many choices are in each row |
---|
72 | for (i = 0; i < n+1; i++) starts[i] = 0; |
---|
73 | for (i = 0; i < nm; i++) { |
---|
74 | traverse_mtbdd_vect_rec(ddman, matrices_bdds[i], rvars, num_vars, 0, odd, 0, 1); |
---|
75 | } |
---|
76 | // and use this to compute the starts information |
---|
77 | // (and at same time, compute max num choices in a state) |
---|
78 | max = 0; |
---|
79 | for (i = 1 ; i < n+1; i++) { |
---|
80 | if (starts[i] > max) max = starts[i]; |
---|
81 | starts[i] += starts[i-1]; |
---|
82 | } |
---|
83 | ndsm->k = max; |
---|
84 | |
---|
85 | // now traverse mtbdds to compute how many transitions in each choice |
---|
86 | for (i = 0; i < nc+1; i++) starts2[i] = 0; |
---|
87 | for (i = 0; i < nm; i++) { |
---|
88 | traverse_mtbdd_matr_rec(ddman, matrices[i], rvars, cvars, num_vars, 0, odd, odd, 0, 0, 10, false); |
---|
89 | traverse_mtbdd_vect_rec(ddman, matrices_bdds[i], rvars, num_vars, 0, odd, 0, 2); |
---|
90 | } |
---|
91 | // and use this to compute the starts2 information |
---|
92 | // (and at same time, compute max num transitions in a choice) |
---|
93 | max2 = 0; |
---|
94 | for (i = 1; i < nc+1; i++) { |
---|
95 | if (starts2[i] > max2) max2 = starts2[i]; |
---|
96 | starts2[i] += starts2[i-1]; |
---|
97 | } |
---|
98 | // recompute starts (because we altered them during last traversal) |
---|
99 | for (i = n; i > 0; i--) { |
---|
100 | starts[i] = starts[i-1]; |
---|
101 | } |
---|
102 | starts[0] = 0; |
---|
103 | |
---|
104 | // max num choices/transitions determines whether we store counts or starts: |
---|
105 | ndsm->use_counts = (max < (unsigned int)(1 << (8*sizeof(unsigned char)))); |
---|
106 | ndsm->use_counts &= (max2 < (unsigned int)(1 << (8*sizeof(unsigned char)))); |
---|
107 | |
---|
108 | // now traverse the mtbdd again to get the actual matrix entries |
---|
109 | for (i = 0; i < nm; i++) { |
---|
110 | traverse_mtbdd_matr_rec(ddman, matrices[i], rvars, cvars, num_vars, 0, odd, odd, 0, 0, 11, false); |
---|
111 | traverse_mtbdd_vect_rec(ddman, matrices_bdds[i], rvars, num_vars, 0, odd, 0, 2); |
---|
112 | } |
---|
113 | // recompute starts (because we altered them during last traversal) |
---|
114 | for (i = n; i > 0; i--) { |
---|
115 | starts[i] = starts[i-1]; |
---|
116 | } |
---|
117 | starts[0] = 0; |
---|
118 | // recompute starts2 (likewise) |
---|
119 | for (i = nc; i > 0; i--) { |
---|
120 | starts2[i] = starts2[i-1]; |
---|
121 | } |
---|
122 | starts2[0] = 0; |
---|
123 | |
---|
124 | // if it's safe to do so, replace starts/starts2 with (smaller) arrays of counts |
---|
125 | if (ndsm->use_counts) { |
---|
126 | ndsm->row_counts = new unsigned char[n]; |
---|
127 | for (i = 0; i < n; i++) ndsm->row_counts[i] = (unsigned char)(starts[i+1] - starts[i]); |
---|
128 | delete[] starts; starts = NULL; |
---|
129 | ndsm->choice_counts = new unsigned char[nc]; |
---|
130 | for (i = 0; i < nc; i++) ndsm->choice_counts[i] = (unsigned char)(starts2[i+1] - starts2[i]); |
---|
131 | delete[] starts2; starts2 = NULL; |
---|
132 | ndsm->mem = (nnz * (sizeof(double) + sizeof(unsigned int)) + (n+nc) * sizeof(unsigned char)) / 1024.0; |
---|
133 | } else { |
---|
134 | ndsm->row_counts = (unsigned char*)starts; |
---|
135 | ndsm->choice_counts = (unsigned char*)starts2; |
---|
136 | ndsm->mem = (nnz * (sizeof(double) + sizeof(unsigned int)) + (n+nc) * sizeof(int)) / 1024.0; |
---|
137 | } |
---|
138 | |
---|
139 | // try/catch for memory allocation/deallocation |
---|
140 | } catch(std::bad_alloc e) { |
---|
141 | if (ndsm) delete ndsm; |
---|
142 | if (matrices) delete[] matrices; |
---|
143 | if (matrices_bdds) { |
---|
144 | for (i = 0; i < nm; i++) Cudd_RecursiveDeref(ddman, matrices_bdds[i]); |
---|
145 | delete[] matrices_bdds; |
---|
146 | } |
---|
147 | if (starts) delete[] starts; |
---|
148 | if (starts2) delete[] starts2; |
---|
149 | throw e; |
---|
150 | } |
---|
151 | |
---|
152 | // clear up memory |
---|
153 | for (i = 0; i < nm; i++) { |
---|
154 | Cudd_RecursiveDeref(ddman, matrices_bdds[i]); |
---|
155 | // nb: don't deref matrices array because that was just pointers, not new copies |
---|
156 | } |
---|
157 | delete[] matrices; |
---|
158 | delete[] matrices_bdds; |
---|
159 | |
---|
160 | return ndsm; |
---|
161 | } |
---|
162 | |
---|
163 | //----------------------------------------------------------------------------------- |
---|
164 | |
---|
165 | // Build nondeterministic (MDP) sparse matrix for "sub-MDP". |
---|
166 | // This function basically exists to construct a sparse matrix representing the transition rewards for an MDP. |
---|
167 | // The complication is that we need to use the nondeterministic choice indexing of the main |
---|
168 | // MDP matrix, not the rewards matrix, otherwise we can't tell which reward is on which transition. |
---|
169 | // throws std::bad_alloc on out-of-memory |
---|
170 | |
---|
171 | NDSparseMatrix *build_sub_nd_sparse_matrix(DdManager *ddman, DdNode *mdp, DdNode *submdp, DdNode **rvars, DdNode **cvars, int num_vars, DdNode **ndvars, int num_ndvars, ODDNode *odd) |
---|
172 | { |
---|
173 | int i, n, nm, nc, nnz, max, max2; |
---|
174 | DdNode *tmp = NULL, **matrices = NULL, **submatrices = NULL, **matrices_bdds = NULL; |
---|
175 | |
---|
176 | // try/catch for memory allocation/deallocation |
---|
177 | try { |
---|
178 | |
---|
179 | // create new data structure |
---|
180 | ndsm = NULL; ndsm = new NDSparseMatrix(); |
---|
181 | |
---|
182 | // get number of states from odd |
---|
183 | n = ndsm->n = odd->eoff+odd->toff; |
---|
184 | // get num of choices (prob. distributions) (USING MDP) |
---|
185 | Cudd_Ref(mdp); |
---|
186 | tmp = DD_ThereExists(ddman, DD_Not(ddman, DD_Equals(ddman, mdp, 0)), cvars, num_vars); |
---|
187 | nc = ndsm->nc = (int)DD_GetNumMinterms(ddman, tmp, num_vars+num_ndvars); |
---|
188 | // get num of transitions (USING SUB-MDP) |
---|
189 | nnz = ndsm->nnz = (int)DD_GetNumMinterms(ddman, submdp, num_vars*2+num_ndvars); |
---|
190 | // break the two mtbdds (MDP AND SUB-MDP) into several (nm) mtbdds |
---|
191 | tmp = DD_ThereExists(ddman, tmp, rvars, num_vars); |
---|
192 | nm = (int)DD_GetNumMinterms(ddman, tmp, num_ndvars); |
---|
193 | Cudd_RecursiveDeref(ddman, tmp); |
---|
194 | matrices = new DdNode*[nm]; |
---|
195 | submatrices = new DdNode*[nm]; |
---|
196 | count = 0; |
---|
197 | split_mdp_and_sub_mdp_rec(ddman, mdp, submdp, ndvars, num_ndvars, 0, matrices, submatrices); |
---|
198 | // and for each one create a bdd storing which rows/choices are non-empty (USING MDP) |
---|
199 | matrices_bdds = new DdNode*[nm]; |
---|
200 | for (i = 0; i < nm; i++) { |
---|
201 | Cudd_Ref(matrices[i]); |
---|
202 | matrices_bdds[i] = DD_ThereExists(ddman, DD_Not(ddman, DD_Equals(ddman, matrices[i], 0)), cvars, num_vars); |
---|
203 | } |
---|
204 | |
---|
205 | // create arrays |
---|
206 | ndsm->non_zeros = new double[nnz]; |
---|
207 | ndsm->cols = new unsigned int[nnz]; |
---|
208 | starts = NULL; starts = new int[n+1]; |
---|
209 | starts2 = NULL; starts2 = new int[nc+1]; |
---|
210 | |
---|
211 | // first traverse mtbdds to compute how many choices are in each row (USING MDP) |
---|
212 | for (i = 0; i < n+1; i++) starts[i] = 0; |
---|
213 | for (i = 0; i < nm; i++) { |
---|
214 | traverse_mtbdd_vect_rec(ddman, matrices_bdds[i], rvars, num_vars, 0, odd, 0, 1); |
---|
215 | } |
---|
216 | // and use this to compute the starts information |
---|
217 | // (and at same time, compute max num choices in a state) |
---|
218 | max = 0; |
---|
219 | for (i = 1 ; i < n+1; i++) { |
---|
220 | if (starts[i] > max) max = starts[i]; |
---|
221 | starts[i] += starts[i-1]; |
---|
222 | } |
---|
223 | ndsm->k = max; |
---|
224 | |
---|
225 | // now traverse mtbdds to compute how many transitions in each choice (USING SUB-MDP) |
---|
226 | for (i = 0; i < nc+1; i++) starts2[i] = 0; |
---|
227 | for (i = 0; i < nm; i++) { |
---|
228 | traverse_mtbdd_matr_rec(ddman, submatrices[i], rvars, cvars, num_vars, 0, odd, odd, 0, 0, 10, false); |
---|
229 | traverse_mtbdd_vect_rec(ddman, matrices_bdds[i], rvars, num_vars, 0, odd, 0, 2); |
---|
230 | } |
---|
231 | // and use this to compute the starts2 information |
---|
232 | // (and at same time, compute max num transitions in a choice) |
---|
233 | max2 = 0; |
---|
234 | for (i = 1; i < nc+1; i++) { |
---|
235 | if (starts2[i] > max2) max2 = starts2[i]; |
---|
236 | starts2[i] += starts2[i-1]; |
---|
237 | } |
---|
238 | // recompute starts (because we altered them during last traversal) |
---|
239 | for (i = n; i > 0; i--) { |
---|
240 | starts[i] = starts[i-1]; |
---|
241 | } |
---|
242 | starts[0] = 0; |
---|
243 | |
---|
244 | // max num choices/transitions determines whether we store counts or starts: |
---|
245 | ndsm->use_counts = (max < (unsigned int)(1 << (8*sizeof(unsigned char)))); |
---|
246 | ndsm->use_counts &= (max2 < (unsigned int)(1 << (8*sizeof(unsigned char)))); |
---|
247 | |
---|
248 | // now traverse the mtbdd again to get the actual matrix entries (USING SUB-MDP) |
---|
249 | for (i = 0; i < nm; i++) { |
---|
250 | traverse_mtbdd_matr_rec(ddman, submatrices[i], rvars, cvars, num_vars, 0, odd, odd, 0, 0, 11, false); |
---|
251 | traverse_mtbdd_vect_rec(ddman, matrices_bdds[i], rvars, num_vars, 0, odd, 0, 2); |
---|
252 | } |
---|
253 | // recompute starts (because we altered them during last traversal) |
---|
254 | for (i = n; i > 0; i--) { |
---|
255 | starts[i] = starts[i-1]; |
---|
256 | } |
---|
257 | starts[0] = 0; |
---|
258 | // recompute starts2 (likewise) |
---|
259 | for (i = nc; i > 0; i--) { |
---|
260 | starts2[i] = starts2[i-1]; |
---|
261 | } |
---|
262 | starts2[0] = 0; |
---|
263 | |
---|
264 | // if it's safe to do so, replace starts/starts2 with (smaller) arrays of counts |
---|
265 | if (ndsm->use_counts) { |
---|
266 | ndsm->row_counts = new unsigned char[n]; |
---|
267 | for (i = 0; i < n; i++) ndsm->row_counts[i] = (unsigned char)(starts[i+1] - starts[i]); |
---|
268 | delete[] starts; starts = NULL; |
---|
269 | ndsm->choice_counts = new unsigned char[nc]; |
---|
270 | for (i = 0; i < nc; i++) ndsm->choice_counts[i] = (unsigned char)(starts2[i+1] - starts2[i]); |
---|
271 | delete[] starts2; starts2 = NULL; |
---|
272 | ndsm->mem = (nnz * (sizeof(double) + sizeof(unsigned int)) + (n+nc) * sizeof(unsigned char)) / 1024.0; |
---|
273 | } else { |
---|
274 | ndsm->row_counts = (unsigned char*)starts; |
---|
275 | ndsm->choice_counts = (unsigned char*)starts2; |
---|
276 | ndsm->mem = (nnz * (sizeof(double) + sizeof(unsigned int)) + (n+nc) * sizeof(int)) / 1024.0; |
---|
277 | } |
---|
278 | |
---|
279 | // try/catch for memory allocation/deallocation |
---|
280 | } catch(std::bad_alloc e) { |
---|
281 | if (ndsm) delete ndsm; |
---|
282 | if (matrices) delete[] matrices; |
---|
283 | if (submatrices) delete[] submatrices; |
---|
284 | if (matrices_bdds) { |
---|
285 | for (i = 0; i < nm; i++) Cudd_RecursiveDeref(ddman, matrices_bdds[i]); |
---|
286 | delete[] matrices_bdds; |
---|
287 | } |
---|
288 | if (starts) delete[] starts; |
---|
289 | if (starts2) delete[] starts2; |
---|
290 | throw e; |
---|
291 | } |
---|
292 | |
---|
293 | // clear up memory |
---|
294 | for (i = 0; i < nm; i++) { |
---|
295 | Cudd_RecursiveDeref(ddman, matrices_bdds[i]); |
---|
296 | // nb: don't deref matrices/submatrices array because that was just pointers, not new copies |
---|
297 | } |
---|
298 | delete[] matrices; |
---|
299 | delete[] submatrices; |
---|
300 | delete[] matrices_bdds; |
---|
301 | |
---|
302 | return ndsm; |
---|
303 | } |
---|
304 | |
---|
305 | //------------------------------------------------------------------------------ |
---|
306 | |
---|
307 | // traverses the mtbdd and gets all the MATRIX entries out |
---|
308 | // does different things depending on the value of 'code' |
---|
309 | // if tranpose flag is true, actually extract from tranpose of matrix |
---|
310 | |
---|
311 | void traverse_mtbdd_matr_rec(DdManager *ddman, DdNode *dd, DdNode **rvars, DdNode **cvars, int num_vars, int level, ODDNode *row, ODDNode *col, int r, int c, int code, bool transpose) |
---|
312 | { |
---|
313 | DdNode *e, *t, *ee, *et, *te, *tt; |
---|
314 | int i, dist_num; |
---|
315 | double *dist, d; |
---|
316 | |
---|
317 | // base case - zero terminal |
---|
318 | if (dd == Cudd_ReadZero(ddman)) return; |
---|
319 | |
---|
320 | // base case - non zero terminal |
---|
321 | if (level == num_vars) { |
---|
322 | switch (code) { |
---|
323 | |
---|
324 | // row major - first pass |
---|
325 | case 1: |
---|
326 | starts[(transpose?c:r)+1]++; |
---|
327 | break; |
---|
328 | |
---|
329 | // row major - second pass |
---|
330 | case 2: |
---|
331 | rmsm->non_zeros[starts[(transpose?c:r)]] = Cudd_V(dd); |
---|
332 | rmsm->cols[starts[(transpose?c:r)]] = (transpose?r:c); |
---|
333 | starts[(transpose?c:r)]++; |
---|
334 | break; |
---|
335 | |
---|
336 | // column major - first pass |
---|
337 | case 3: |
---|
338 | starts[(transpose?r:c)+1]++; |
---|
339 | break; |
---|
340 | |
---|
341 | // column major - second pass |
---|
342 | case 4: |
---|
343 | cmsm->non_zeros[starts[(transpose?r:c)]] = Cudd_V(dd); |
---|
344 | cmsm->rows[starts[(transpose?r:c)]] = (transpose?c:r); |
---|
345 | starts[(transpose?r:c)]++; |
---|
346 | break; |
---|
347 | |
---|
348 | // row/column - only pass |
---|
349 | case 5: |
---|
350 | rcsm->non_zeros[count] = Cudd_V(dd); |
---|
351 | rcsm->rows[count] = (transpose?c:r); |
---|
352 | rcsm->cols[count] = (transpose?r:c); |
---|
353 | count++; |
---|
354 | break; |
---|
355 | |
---|
356 | // compact modified sparse row - first pass |
---|
357 | case 6: |
---|
358 | starts[(transpose?c:r)+1]++; |
---|
359 | break; |
---|
360 | |
---|
361 | // compact modified sparse row - second pass |
---|
362 | case 7: |
---|
363 | // try and find value |
---|
364 | dist = cmsrsm->dist; |
---|
365 | dist_num = cmsrsm->dist_num; |
---|
366 | d = Cudd_V(dd); |
---|
367 | for (i = 0; i < dist_num; i++) if (dist[i] == d) break; |
---|
368 | // if it's not there, add it |
---|
369 | if (i == dist_num) { |
---|
370 | dist[i] = d; |
---|
371 | cmsrsm->dist_num++; |
---|
372 | } |
---|
373 | // store info |
---|
374 | cmsrsm->cols[starts[(transpose?c:r)]] = (unsigned int)(((unsigned int)(transpose?r:c) << cmsrsm->dist_shift) + (unsigned int)i); |
---|
375 | starts[(transpose?c:r)]++; |
---|
376 | break; |
---|
377 | |
---|
378 | // compact modified sparse column - first pass |
---|
379 | case 8: |
---|
380 | starts[(transpose?r:c)+1]++; |
---|
381 | break; |
---|
382 | |
---|
383 | // compact modified sparse column - second pass |
---|
384 | case 9: |
---|
385 | // try and find value |
---|
386 | dist = cmscsm->dist; |
---|
387 | dist_num = cmscsm->dist_num; |
---|
388 | d = Cudd_V(dd); |
---|
389 | for (i = 0; i < dist_num; i++) if (dist[i] == d) break; |
---|
390 | // if it's not there, add it |
---|
391 | if (i == dist_num) { |
---|
392 | dist[i] = d; |
---|
393 | cmscsm->dist_num++; |
---|
394 | } |
---|
395 | // store info |
---|
396 | cmscsm->rows[starts[(transpose?r:c)]] = (unsigned int)(((unsigned int)(transpose?c:r) << cmscsm->dist_shift) + (unsigned int)i); |
---|
397 | starts[(transpose?r:c)]++; |
---|
398 | break; |
---|
399 | |
---|
400 | // mdp - first pass |
---|
401 | case 10: |
---|
402 | starts2[starts[(transpose?c:r)]+1]++; |
---|
403 | break; |
---|
404 | |
---|
405 | // mdp - second pass |
---|
406 | case 11: |
---|
407 | ndsm->non_zeros[starts2[starts[(transpose?c:r)]]] = Cudd_V(dd); |
---|
408 | ndsm->cols[starts2[starts[(transpose?c:r)]]] = (transpose?r:c); |
---|
409 | starts2[starts[(transpose?c:r)]]++; |
---|
410 | break; |
---|
411 | } |
---|
412 | return; |
---|
413 | } |
---|
414 | |
---|
415 | // recurse |
---|
416 | if (dd->index > cvars[level]->index) { |
---|
417 | ee = et = te = tt = dd; |
---|
418 | } |
---|
419 | else if (dd->index > rvars[level]->index) { |
---|
420 | ee = te = Cudd_E(dd); |
---|
421 | et = tt = Cudd_T(dd); |
---|
422 | } |
---|
423 | else { |
---|
424 | e = Cudd_E(dd); |
---|
425 | if (e->index > cvars[level]->index) { |
---|
426 | ee = et = e; |
---|
427 | } |
---|
428 | else { |
---|
429 | ee = Cudd_E(e); |
---|
430 | et = Cudd_T(e); |
---|
431 | } |
---|
432 | t = Cudd_T(dd); |
---|
433 | if (t->index > cvars[level]->index) { |
---|
434 | te = tt = t; |
---|
435 | } |
---|
436 | else { |
---|
437 | te = Cudd_E(t); |
---|
438 | tt = Cudd_T(t); |
---|
439 | } |
---|
440 | } |
---|
441 | |
---|
442 | traverse_mtbdd_matr_rec(ddman, ee, rvars, cvars, num_vars, level+1, row->e, col->e, r, c, code, transpose); |
---|
443 | traverse_mtbdd_matr_rec(ddman, et, rvars, cvars, num_vars, level+1, row->e, col->t, r, c+col->eoff, code, transpose); |
---|
444 | traverse_mtbdd_matr_rec(ddman, te, rvars, cvars, num_vars, level+1, row->t, col->e, r+row->eoff, c, code, transpose); |
---|
445 | traverse_mtbdd_matr_rec(ddman, tt, rvars, cvars, num_vars, level+1, row->t, col->t, r+row->eoff, c+col->eoff, code, transpose); |
---|
446 | } |
---|
447 | |
---|