12008-07-03  Brian Gough  <bjg@hp2.network-theory.co.uk>
2
3	* Makefile.am (INCLUDES): use top_srcdir instead of top_builddir
4
52007-05-30  Brian Gough  <bjg@network-theory.co.uk>
6
7	* siman.c (boltzmann): moved safe_exp into boltzmann
8	(copy_state): put all copying into one function
9
102006-03-08  Brian Gough  <bjg@network-theory.co.uk>
11
12	* test.c (square): removed inline since it causes problems with
13	some compilers
14
152005-11-14  Brian Gough  <bjg@network-theory.co.uk>
16
17	* siman.c (safe_exp): added a safe_exp function to avoid underflow
18	for large uphill steps
19
202003-03-31  Brian Gough  <bjg@network-theory.co.uk>
21
22	* siman.c (gsl_siman_solve): avoid reevaluation for best_E
23	(gsl_siman_solve): loop over param.iters_fixed_T not
24	params.n_tries
25	(gsl_siman_solve): initialise energy at start
26
27Sat Aug  3 20:32:38 2002  Brian Gough  <bjg@network-theory.co.uk>
28
29	* siman.c (gsl_siman_solve): fix acceptance criterion to match
30	documentation (Peter S. Christopher)
31
32Thu Jun 13 20:57:00 2002  Brian Gough  <bjg@network-theory.co.uk>
33
34	* siman.c (gsl_siman_solve): keep track of the best result
35
362002-02-07  Mark Galassi  <rosalia@galassi.org>
37
38	* siman.c (gsl_siman_solve): bug fix in the destructor for x and
39	new_x which was being called on &x and &new_x instead of x and
40	new_x; thanks to Karsten Howes <karsten@videotron.ca>
41
42Thu Jul 12 21:50:07 2001  Brian Gough  <bjg@network-theory.co.uk>
43
44	* gsl_siman.h: changed renamed gsl_Efunc_t to gsl_siman_Efunc_t,
45	in accordance with namespace conventions
46
472000-12-15  Mark Galassi  <rosalia@galassi.org>
48
49	* siman.c (gsl_siman_solve): reversed a small change I had made
50	earlier and went back to taking Boltzmann-conditional steps when
51	the new energy is equal to the previous one.  This allows you to
52	move around if you are stuck on a plateau.
53
54	* gsl_siman.h, siman.c, siman_test.c, siman_tsp.c, test.c: changed
55	the siman_solve API to allow for more general search spaces.  The
56	problem was that we assumed that points in the search space were
57	data structures that were allocated in continguous memory, so they
58	could not be linked structures.  I replaced the malloc(), memcpy()
59	and free() calls with copy_constructor(), copyfunc() and
60	copy_destructor() functions.  The user passes these functions,
61	which means that siman_solve() now takes three more arguments of
62	type gsl_siman_copy_t, gsl_siman_copy_construct_t and
63	gsl_siman_destroy_t.  If these arguments are NULL (and all three
64	of them have to be NULL together), the traditional memcpy()
65	approach is used.
66
671999-02-14  Mark Galassi  <rosalia@cygnus.com>
68
69	* minor fixes.
70
71Tue Nov 17 17:22:14 1998  Brian Gough  <bjg@vvv.lanl.gov>
72
73	* added #include <config.h> to all top-level source files
74
75Sun Nov  8 20:40:28 1998  Brian Gough  <bjg@vvv.lanl.gov>
76
77	* siman_tsp.c: clean up for make strict
78
791998-11-06    <bjg@ancho.lanl.gov>
80
81	* test.c: added prototype for memcpy using #include <string.h>
82
83	* siman_test.c: added prototype for memcpy using #include <string.h>
84
85Wed Oct 28 15:06:58 1998  Brian Gough  <bjg@vvv.lanl.gov>
86
87	* siman.c: added #include <string.h> for memcpy
88
89Thu Aug 20 12:22:28 1998  Brian Gough  <bjg@vvv.lanl.gov>
90
91	* siman.c: use (char *) judiciously to avoid warnings about void
92 	pointer arithmetic (see randist/shuffle.c for similar examples)
93
94	* siman_test.c: perform several tests, using the exact answer as
95 	the comparison value, rather than checking for stationarity.
96
97Sun Jun 28 14:11:04 1998  Brian Gough  <bjg@vvv.lanl.gov>
98
99	* Converted to work with rng-style random number generators
100
101	* gsl_siman.h: gsl_siman_step_t type functions now take a gsl_rng
102 	random number generator as their first argument
103
104	* siman.c (gsl_siman_solve): Now takes a gsl_rng random number
105 	generator as the first argument
106
107Fri Jun 19 11:17:24 1998  Brian Gough  <bjg@vvv.lanl.gov>
108
109	* siman.c (gsl_siman_solve_many): changed the variable 'throw' to
110 	'u' (for uniform-random-number) so that we can compile with c++
111 	where throw is a reserved word.
112
113Sat May 23 13:59:55 1998  Brian Gough  <bjg@vvv.lanl.gov>
114
115	* siman.c: made the solving functions deterministic by removing
116 	the random seed, gsl_ran_seed(time(0L)). When the function is
117 	non-deterministic it is hard to debug and test (about 1 time in 20
118 	the test would fail due to the randomness). We can let the user do
119 	the seeding if they need that.
120
1211998-02-09  Mark Galassi  <rosalia@nis.lanl.gov>
122
123	* siman_test_driver.sh (LAST_ENERGY): fixed a typo; the tests now
124	report well when they converge.
125
1261998-01-30  Mark Galassi  <rosalia@cygnus.com>
127
128	* siman_test_driver.sh, Makefile.am (TESTS): added a test driver
129	so that now "make check" does something interesting.
130