The source code and dockerfile for the GSW2024 AI Lab.
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
This repo is archived. You can view files and clone it, but cannot push or open issues/pull-requests.
 
 
 
 
 
 

537 lines
16 KiB

/**
@file
@ingroup cudd
@brief Functions to read in a matrix
@see cudd_addHarwell.c
@author Fabio Somenzi
@copyright@parblock
Copyright (c) 1995-2015, Regents of the University of Colorado
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions
are met:
Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in the
documentation and/or other materials provided with the distribution.
Neither the name of the University of Colorado nor the names of its
contributors may be used to endorse or promote products derived from
this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
POSSIBILITY OF SUCH DAMAGE.
@endparblock
*/
#include "util.h"
#include "cuddInt.h"
/*---------------------------------------------------------------------------*/
/* Constant declarations */
/*---------------------------------------------------------------------------*/
/*---------------------------------------------------------------------------*/
/* Stucture declarations */
/*---------------------------------------------------------------------------*/
/*---------------------------------------------------------------------------*/
/* Type declarations */
/*---------------------------------------------------------------------------*/
/*---------------------------------------------------------------------------*/
/* Variable declarations */
/*---------------------------------------------------------------------------*/
/*---------------------------------------------------------------------------*/
/* Macro declarations */
/*---------------------------------------------------------------------------*/
/** \cond */
/*---------------------------------------------------------------------------*/
/* Static function prototypes */
/*---------------------------------------------------------------------------*/
/** \endcond */
/*---------------------------------------------------------------------------*/
/* Definition of exported functions */
/*---------------------------------------------------------------------------*/
/**
@brief Reads in a sparse matrix.
@details Reads in a sparse matrix specified in a simple format.
The first line of the input contains the numbers of rows and columns.
The remaining lines contain the elements of the matrix, one per line.
Given a background value
(specified by the background field of the manager), only the values
different from it are explicitly listed. Each foreground element is
described by two integers, i.e., the row and column number, and a
real number, i.e., the value.<p>
Cudd_addRead produces an %ADD that depends on two sets of variables: x
and y. The x variables (x\[0\] ... x\[nx-1\]) encode the row index and
the y variables (y\[0\] ... y\[ny-1\]) encode the column index.
x\[0\] and y\[0\] are the most significant bits in the indices.
The variables may already exist or may be created by the function.
The index of x\[i\] is bx+i*sx, and the index of y\[i\] is by+i*sy.<p>
On input, nx and ny hold the numbers
of row and column variables already in existence. On output, they
hold the numbers of row and column variables actually used by the
matrix. When Cudd_addRead creates the variable arrays,
the index of x\[i\] is bx+i*sx, and the index of y\[i\] is by+i*sy.
When some variables already exist Cudd_addRead expects the indices
of the existing x variables to be bx+i*sx, and the indices of the
existing y variables to be by+i*sy.<p>
m and n are set to the numbers of rows and columns of the
matrix. Their values on input are immaterial.
The %ADD for the sparse matrix is returned in E, and its reference
count is > 0.
@return 1 in case of success; 0 otherwise.
@sideeffect nx and ny are set to the numbers of row and column
variables. m and n are set to the numbers of rows and columns. x and y
are possibly extended to represent the array of row and column
variables. Similarly for xn and yn_, which hold on return from
Cudd_addRead the complements of the row and column variables.
@see Cudd_addHarwell Cudd_bddRead
*/
int
Cudd_addRead(
FILE * fp /**< input file pointer */,
DdManager * dd /**< %DD manager */,
DdNode ** E /**< characteristic function of the graph */,
DdNode *** x /**< array of row variables */,
DdNode *** y /**< array of column variables */,
DdNode *** xn /**< array of complemented row variables */,
DdNode *** yn_ /**< array of complemented column variables */,
int * nx /**< number or row variables */,
int * ny /**< number or column variables */,
int * m /**< number of rows */,
int * n /**< number of columns */,
int bx /**< first index of row variables */,
int sx /**< step of row variables */,
int by /**< first index of column variables */,
int sy /**< step of column variables */)
{
DdNode *one, *zero;
DdNode *w, *neW;
DdNode *minterm1;
int u, v, err, i, nv;
int lnx, lny;
CUDD_VALUE_TYPE val;
DdNode **lx, **ly, **lxn, **lyn;
one = DD_ONE(dd);
zero = DD_ZERO(dd);
err = fscanf(fp, "%d %d", &u, &v);
if (err == EOF) {
return(0);
} else if (err != 2) {
return(0);
}
*m = u;
/* Compute the number of x variables. */
lx = *x; lxn = *xn;
u--; /* row and column numbers start from 0 */
for (lnx=0; u > 0; lnx++) {
u >>= 1;
}
/* Here we rely on the fact that REALLOC of a null pointer is
** translates to an ALLOC.
*/
if (lnx > *nx) {
*x = lx = REALLOC(DdNode *, *x, lnx);
if (lx == NULL) {
dd->errorCode = CUDD_MEMORY_OUT;
return(0);
}
*xn = lxn = REALLOC(DdNode *, *xn, lnx);
if (lxn == NULL) {
dd->errorCode = CUDD_MEMORY_OUT;
return(0);
}
}
*n = v;
/* Compute the number of y variables. */
ly = *y; lyn = *yn_;
v--; /* row and column numbers start from 0 */
for (lny=0; v > 0; lny++) {
v >>= 1;
}
/* Here we rely on the fact that REALLOC of a null pointer is
** translates to an ALLOC.
*/
if (lny > *ny) {
*y = ly = REALLOC(DdNode *, *y, lny);
if (ly == NULL) {
dd->errorCode = CUDD_MEMORY_OUT;
return(0);
}
*yn_ = lyn = REALLOC(DdNode *, *yn_, lny);
if (lyn == NULL) {
dd->errorCode = CUDD_MEMORY_OUT;
return(0);
}
}
/* Create all new variables. */
for (i = *nx, nv = bx + (*nx) * sx; i < lnx; i++, nv += sx) {
do {
dd->reordered = 0;
lx[i] = cuddUniqueInter(dd, nv, one, zero);
} while (dd->reordered == 1);
if (lx[i] == NULL) {
if (dd->errorCode == CUDD_TIMEOUT_EXPIRED && dd->timeoutHandler) {
dd->timeoutHandler(dd, dd->tohArg);
}
return(0);
}
cuddRef(lx[i]);
do {
dd->reordered = 0;
lxn[i] = cuddUniqueInter(dd, nv, zero, one);
} while (dd->reordered == 1);
if (lxn[i] == NULL) {
if (dd->errorCode == CUDD_TIMEOUT_EXPIRED && dd->timeoutHandler) {
dd->timeoutHandler(dd, dd->tohArg);
}
return(0);
}
cuddRef(lxn[i]);
}
for (i = *ny, nv = by + (*ny) * sy; i < lny; i++, nv += sy) {
do {
dd->reordered = 0;
ly[i] = cuddUniqueInter(dd, nv, one, zero);
} while (dd->reordered == 1);
if (ly[i] == NULL) {
if (dd->errorCode == CUDD_TIMEOUT_EXPIRED && dd->timeoutHandler) {
dd->timeoutHandler(dd, dd->tohArg);
}
return(0);
}
cuddRef(ly[i]);
do {
dd->reordered = 0;
lyn[i] = cuddUniqueInter(dd, nv, zero, one);
} while (dd->reordered == 1);
if (lyn[i] == NULL) {
if (dd->errorCode == CUDD_TIMEOUT_EXPIRED && dd->timeoutHandler) {
dd->timeoutHandler(dd, dd->tohArg);
}
return(0);
}
cuddRef(lyn[i]);
}
*nx = lnx;
*ny = lny;
*E = dd->background; /* this call will never cause reordering */
cuddRef(*E);
while (! feof(fp)) {
err = fscanf(fp, "%d %d %lf", &u, &v, &val);
if (err == EOF) {
break;
} else if (err != 3) {
return(0);
} else if (u >= *m || v >= *n || u < 0 || v < 0) {
return(0);
}
minterm1 = one; cuddRef(minterm1);
/* Build minterm1 corresponding to this arc */
for (i = lnx - 1; i>=0; i--) {
if (u & 1) {
w = Cudd_addApply(dd, Cudd_addTimes, minterm1, lx[i]);
} else {
w = Cudd_addApply(dd, Cudd_addTimes, minterm1, lxn[i]);
}
if (w == NULL) {
Cudd_RecursiveDeref(dd, minterm1);
return(0);
}
cuddRef(w);
Cudd_RecursiveDeref(dd, minterm1);
minterm1 = w;
u >>= 1;
}
for (i = lny - 1; i>=0; i--) {
if (v & 1) {
w = Cudd_addApply(dd, Cudd_addTimes, minterm1, ly[i]);
} else {
w = Cudd_addApply(dd, Cudd_addTimes, minterm1, lyn[i]);
}
if (w == NULL) {
Cudd_RecursiveDeref(dd, minterm1);
return(0);
}
cuddRef(w);
Cudd_RecursiveDeref(dd, minterm1);
minterm1 = w;
v >>= 1;
}
/* Create new constant node if necessary.
** This call will never cause reordering.
*/
neW = cuddUniqueConst(dd, val);
if (neW == NULL) {
Cudd_RecursiveDeref(dd, minterm1);
return(0);
}
cuddRef(neW);
w = Cudd_addIte(dd, minterm1, neW, *E);
if (w == NULL) {
Cudd_RecursiveDeref(dd, minterm1);
Cudd_RecursiveDeref(dd, neW);
return(0);
}
cuddRef(w);
Cudd_RecursiveDeref(dd, minterm1);
Cudd_RecursiveDeref(dd, neW);
Cudd_RecursiveDeref(dd, *E);
*E = w;
}
return(1);
} /* end of Cudd_addRead */
/**
@brief Reads in a graph (without labels) given as a list of arcs.
@details Reads in a graph (without labels) given as an adjacency
matrix. The first line of the input contains the numbers of rows and
columns of the adjacency matrix. The remaining lines contain the arcs
of the graph, one per line. Each arc is described by two integers,
i.e., the row and column number, or the indices of the two endpoints.
Cudd_bddRead produces a %BDD that depends on two sets of variables: x
and y. The x variables (x\[0\] ... x\[nx-1\]) encode
the row index and the y variables (y\[0\] ... y\[ny-1\]) encode the
column index. x\[0\] and y\[0\] are the most significant bits in the
indices.
The variables may already exist or may be created by the function.
The index of x\[i\] is bx+i*sx, and the index of y\[i\] is by+i*sy.<p>
On input, nx and ny hold the numbers of row and column variables already
in existence. On output, they hold the numbers of row and column
variables actually used by the matrix. When Cudd_bddRead creates the
variable arrays, the index of x\[i\] is bx+i*sx, and the index of
y\[i\] is by+i*sy. When some variables already exist, Cudd_bddRead
expects the indices of the existing x variables to be bx+i*sx, and the
indices of the existing y variables to be by+i*sy.<p>
m and n are set to the numbers of rows and columns of the
matrix. Their values on input are immaterial. The %BDD for the graph
is returned in E, and its reference count is > 0.
@return 1 in case of success; 0 otherwise.
@sideeffect nx and ny are set to the numbers of row and column
variables. m and n are set to the numbers of rows and columns. x and y
are possibly extended to represent the array of row and column
variables.
@see Cudd_addHarwell Cudd_addRead
*/
int
Cudd_bddRead(
FILE * fp /**< input file pointer */,
DdManager * dd /**< DD manager */,
DdNode ** E /**< characteristic function of the graph */,
DdNode *** x /**< array of row variables */,
DdNode *** y /**< array of column variables */,
int * nx /**< number or row variables */,
int * ny /**< number or column variables */,
int * m /**< number of rows */,
int * n /**< number of columns */,
int bx /**< first index of row variables */,
int sx /**< step of row variables */,
int by /**< first index of column variables */,
int sy /**< step of column variables */)
{
DdNode *one, *zero;
DdNode *w;
DdNode *minterm1;
int u, v, err, i, nv;
int lnx, lny;
DdNode **lx, **ly;
one = DD_ONE(dd);
zero = Cudd_Not(one);
err = fscanf(fp, "%d %d", &u, &v);
if (err == EOF) {
return(0);
} else if (err != 2) {
return(0);
}
*m = u;
/* Compute the number of x variables. */
lx = *x;
u--; /* row and column numbers start from 0 */
for (lnx=0; u > 0; lnx++) {
u >>= 1;
}
if (lnx > *nx) {
*x = lx = REALLOC(DdNode *, *x, lnx);
if (lx == NULL) {
dd->errorCode = CUDD_MEMORY_OUT;
return(0);
}
}
*n = v;
/* Compute the number of y variables. */
ly = *y;
v--; /* row and column numbers start from 0 */
for (lny=0; v > 0; lny++) {
v >>= 1;
}
if (lny > *ny) {
*y = ly = REALLOC(DdNode *, *y, lny);
if (ly == NULL) {
dd->errorCode = CUDD_MEMORY_OUT;
return(0);
}
}
/* Create all new variables. */
for (i = *nx, nv = bx + (*nx) * sx; i < lnx; i++, nv += sx) {
do {
dd->reordered = 0;
lx[i] = cuddUniqueInter(dd, nv, one, zero);
} while (dd->reordered == 1);
if (lx[i] == NULL) {
if (dd->errorCode == CUDD_TIMEOUT_EXPIRED && dd->timeoutHandler) {
dd->timeoutHandler(dd, dd->tohArg);
}
return(0);
}
cuddRef(lx[i]);
}
for (i = *ny, nv = by + (*ny) * sy; i < lny; i++, nv += sy) {
do {
dd->reordered = 0;
ly[i] = cuddUniqueInter(dd, nv, one, zero);
} while (dd->reordered == 1);
if (ly[i] == NULL) {
if (dd->errorCode == CUDD_TIMEOUT_EXPIRED && dd->timeoutHandler) {
dd->timeoutHandler(dd, dd->tohArg);
}
return(0);
}
cuddRef(ly[i]);
}
*nx = lnx;
*ny = lny;
*E = zero; /* this call will never cause reordering */
cuddRef(*E);
while (! feof(fp)) {
err = fscanf(fp, "%d %d", &u, &v);
if (err == EOF) {
break;
} else if (err != 2) {
return(0);
} else if (u >= *m || v >= *n || u < 0 || v < 0) {
return(0);
}
minterm1 = one; cuddRef(minterm1);
/* Build minterm1 corresponding to this arc. */
for (i = lnx - 1; i>=0; i--) {
if (u & 1) {
w = Cudd_bddAnd(dd, minterm1, lx[i]);
} else {
w = Cudd_bddAnd(dd, minterm1, Cudd_Not(lx[i]));
}
if (w == NULL) {
Cudd_RecursiveDeref(dd, minterm1);
return(0);
}
cuddRef(w);
Cudd_RecursiveDeref(dd,minterm1);
minterm1 = w;
u >>= 1;
}
for (i = lny - 1; i>=0; i--) {
if (v & 1) {
w = Cudd_bddAnd(dd, minterm1, ly[i]);
} else {
w = Cudd_bddAnd(dd, minterm1, Cudd_Not(ly[i]));
}
if (w == NULL) {
Cudd_RecursiveDeref(dd, minterm1);
return(0);
}
cuddRef(w);
Cudd_RecursiveDeref(dd, minterm1);
minterm1 = w;
v >>= 1;
}
w = Cudd_bddAnd(dd, Cudd_Not(minterm1), Cudd_Not(*E));
if (w == NULL) {
Cudd_RecursiveDeref(dd, minterm1);
return(0);
}
w = Cudd_Not(w);
cuddRef(w);
Cudd_RecursiveDeref(dd, minterm1);
Cudd_RecursiveDeref(dd, *E);
*E = w;
}
return(1);
} /* end of Cudd_bddRead */
/*---------------------------------------------------------------------------*/
/* Definition of internal functions */
/*---------------------------------------------------------------------------*/
/*---------------------------------------------------------------------------*/
/* Definition of static functions */
/*---------------------------------------------------------------------------*/