3 * Collection of all flags used through the GiNaC framework. */
6 * GiNaC Copyright (C) 1999-2003 Johannes Gutenberg University Mainz, Germany
8 * This program is free software; you can redistribute it and/or modify
9 * it under the terms of the GNU General Public License as published by
10 * the Free Software Foundation; either version 2 of the License, or
11 * (at your option) any later version.
13 * This program is distributed in the hope that it will be useful,
14 * but WITHOUT ANY WARRANTY; without even the implied warranty of
15 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 * GNU General Public License for more details.
18 * You should have received a copy of the GNU General Public License
19 * along with this program; if not, write to the Free Software
20 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
23 #ifndef __GINAC_FLAGS_H__
24 #define __GINAC_FLAGS_H__
28 /** Flags to control the behavior of expand(). */
29 class expand_options {
32 expand_indexed = 0x0001, ///< expands (a+b).i to a.i+b.i
33 expand_function_args = 0x0002 ///< expands the arguments of functions
37 /** Flags to control the behavior of subs(). */
41 no_pattern = 0x0001, ///< disable pattern matching
42 subs_no_pattern = 0x0001, // for backwards compatibility
43 algebraic = 0x0002, ///< enable algebraic substitutions
44 subs_algebraic = 0x0002, // for backwards compatibility
45 pattern_is_product = 0x0004, ///< used internally by expairseq::subschildren()
46 pattern_is_not_product = 0x0008 ///< used internally by expairseq::subschildren()
50 /** Flags to control series expansion. */
51 class series_options {
54 /** Suppress branch cuts in series expansion. Branch cuts manifest
55 * themselves as step functions, if this option is not passed. If
56 * it is passed and expansion at a point on a cut is performed, then
57 * the analytic continuation of the function is expanded. */
58 suppress_branchcut = 0x0001
62 /** Switch to control algorithm for determinant computation. */
63 class determinant_algo {
66 /** Let the system choose. A heuristics is applied for automatic
67 * determination of a suitable algorithm. */
69 /** Gauss elimination. If \f$m_{i,j}^{(0)}\f$ are the entries of the
70 * original matrix, then the matrix is transformed into triangular
71 * form by applying the rules
73 * m_{i,j}^{(k+1)} = m_{i,j}^{(k)} - m_{i,k}^{(k)} m_{k,j}^{(k)} / m_{k,k}^{(k)}
75 * The determinant is then just the product of diagonal elements.
76 * Choose this algorithm only for purely numerical matrices. */
78 /** Division-free elimination. This is a modification of Gauss
79 * elimination where the division by the pivot element is not
80 * carried out. If \f$m_{i,j}^{(0)}\f$ are the entries of the
81 * original matrix, then the matrix is transformed into triangular
82 * form by applying the rules
84 * m_{i,j}^{(k+1)} = m_{i,j}^{(k)} m_{k,k}^{(k)} - m_{i,k}^{(k)} m_{k,j}^{(k)}
86 * The determinant can later be computed by inspecting the diagonal
87 * elements only. This algorithm is only there for the purpose of
88 * cross-checks. It is never fast. */
90 /** Laplace elimination. This is plain recursive elimination along
91 * minors although multiple minors are avoided by the algorithm.
92 * Although the algorithm is exponential in complexity it is
93 * frequently the fastest one when the matrix is populated by
94 * complicated symbolic expressions. */
96 /** Bareiss fraction-free elimination. This is a modification of
97 * Gauss elimination where the division by the pivot element is
98 * <EM>delayed</EM> until it can be carried out without computing
99 * GCDs. If \f$m_{i,j}^{(0)}\f$ are the entries of the original
100 * matrix, then the matrix is transformed into triangular form by
103 * m_{i,j}^{(k+1)} = (m_{i,j}^{(k)} m_{k,k}^{(k)} - m_{i,k}^{(k)} m_{k,j}^{(k)}) / m_{k-1,k-1}^{(k-1)}
105 * (We have set \f$m_{-1,-1}^{(-1)}=1\f$ in order to avoid a case
106 * distinction in above formula.) It can be shown that nothing more
107 * than polynomial long division is needed for carrying out the
108 * division. The determinant can then be read of from the lower
109 * right entry. This algorithm is rarely fast for computing
115 /** Switch to control algorithm for linear system solving. */
119 /** Let the system choose. A heuristics is applied for automatic
120 * determination of a suitable algorithm. */
122 /** Gauss elimination. If \f$m_{i,j}^{(0)}\f$ are the entries of the
123 * original matrix, then the matrix is transformed into triangular
124 * form by applying the rules
126 * m_{i,j}^{(k+1)} = m_{i,j}^{(k)} - m_{i,k}^{(k)} m_{k,j}^{(k)} / m_{k,k}^{(k)}
128 * This algorithm is well-suited for numerical matrices but generally
129 * suffers from the expensive division (and computation of GCDs) at
132 /** Division-free elimination. This is a modification of Gauss
133 * elimination where the division by the pivot element is not
134 * carried out. If \f$m_{i,j}^{(0)}\f$ are the entries of the
135 * original matrix, then the matrix is transformed into triangular
136 * form by applying the rules
138 * m_{i,j}^{(k+1)} = m_{i,j}^{(k)} m_{k,k}^{(k)} - m_{i,k}^{(k)} m_{k,j}^{(k)}
140 * This algorithm is only there for the purpose of cross-checks.
141 * It suffers from exponential intermediate expression swell. Use it
142 * only for small systems. */
144 /** Bareiss fraction-free elimination. This is a modification of
145 * Gauss elimination where the division by the pivot element is
146 * <EM>delayed</EM> until it can be carried out without computing
147 * GCDs. If \f$m_{i,j}^{(0)}\f$ are the entries of the original
148 * matrix, then the matrix is transformed into triangular form by
151 * m_{i,j}^{(k+1)} = (m_{i,j}^{(k)} m_{k,k}^{(k)} - m_{i,k}^{(k)} m_{k,j}^{(k)}) / m_{k-1,k-1}^{(k-1)}
153 * (We have set \f$m_{-1,-1}^{(-1)}=1\f$ in order to avoid a case
154 * distinction in above formula.) It can be shown that nothing more
155 * than polynomial long division is needed for carrying out the
156 * division. This is generally the fastest algorithm for solving
157 * linear systems. In contrast to division-free elimination it only
158 * has a linear expression swell. For two-dimensional systems, the
159 * two algorithms are equivalent, however. */
164 /** Flags to store information about the state of an object.
165 * @see basic::flags */
169 dynallocated = 0x0001, ///< heap-allocated (i.e. created by new if we want to be clever and bypass the stack, @see ex::construct_from_basic() )
170 evaluated = 0x0002, ///< .eval() has already done its job
171 expanded = 0x0004, ///< .expand(0) has already done its job (other expand() options ignore this flag)
172 hash_calculated = 0x0008, ///< .calchash() has already done its job
173 not_shareable = 0x0010 ///< don't share instances of this object between different expressions unless explicitly asked to (used by ex::compare())
177 /** Possible attributes an object can have. */
181 // answered by class numeric
198 // answered by class relation
203 relation_less_or_equal,
205 relation_greater_or_equal,
207 // answered by class symbol
210 // answered by class lst
213 // answered by class exprseq
216 // answered by classes numeric, symbol, add, mul, power
221 crational_polynomial,
225 // answered by class indexed
226 indexed, // class can carry indices
227 has_indices, // object has at least one index
229 // answered by class idx
239 noncommutative_composite
243 /** Strategies how to clean up the function remember cache.
244 * @see remember_table */
245 class remember_strategies {
248 delete_never, ///< Let table grow undefinitely
249 delete_lru, ///< Least recently used
250 delete_lfu, ///< Least frequently used
251 delete_cyclic ///< First (oldest) one in list
257 #endif // ndef __GINAC_FLAGS_H__