1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
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
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
|
2009-07-25 Brian Gough <bjg@network-theory.co.uk>
* vegas.c (gsl_monte_vegas_params_get)
(gsl_monte_vegas_params_set): added get/set functions for params
(gsl_monte_vegas_chisq): added chisq accessor
* gsl_monte_vegas.h: added separate params struct
* miser.c (gsl_monte_miser_params_set)
(gsl_monte_miser_params_get): added get/set functions for params
* gsl_monte_miser.h: added separate params struct
2009-07-09 Brian Gough <bjg@network-theory.co.uk>
* vegas.c (gsl_monte_vegas_free): handle NULL argument in free
* plain.c (gsl_monte_plain_free): handle NULL argument in free
* miser.c (gsl_monte_miser_free): handle NULL argument in free
2009-02-10 Brian Gough <bjg@network-theory.co.uk>
* vegas.c (gsl_monte_vegas_integrate): use gsl_pow_int to compute
tot_boxes, avoids potentially inaccurate pow functions (MinGW).
2008-11-20 Brian Gough <bjg@network-theory.co.uk>
* vegas.c (gsl_monte_vegas_integrate): improve the chisq
calculation to avoid cancellation errors in the original
formula (fixes bug #24510)
* test.c (MONTE_ERROR_TEST): added test cases for negative chisq
in vegas
2008-07-03 Brian Gough <bjg@hp2.network-theory.co.uk>
* Makefile.am (INCLUDES): use top_srcdir instead of top_builddir
2008-04-26 Brian Gough <bjg@network-theory.co.uk>
* vegas.c (gsl_monte_vegas_integrate): compute running totals as
volatile double to prevent problems with excess precision, use
meaningful variable names for sum of squares, variance and sigma.
* test_main.c: compute the ensemble mean more accurately for the tests
* test.c: added a test for warm-start vegas, as well as cold-start
* miser.c (gsl_monte_miser_integrate): catch zero weights to avoid
division by zero
* test.c (main): added test for step-type function
2004-06-02 Brian Gough <bjg@network-theory.co.uk>
* test_main.c: handle the case where sd==0 && error[i] !=0 more
carefully
Mon Apr 23 13:23:47 2001 Brian Gough <bjg@network-theory.co.uk>
* test_main.c: removed unused status variable
Sat Jan 6 19:56:49 2001 Brian Gough <bjg@network-theory.co.uk>
* miser.c (gsl_monte_miser_free): fixed memory leak for s->hits_{r,l}
* vegas.c (gsl_monte_vegas_free): fixed memory leak for s->x
Fri Dec 22 21:43:04 2000 Brian Gough <bjg@network-theory.co.uk>
* miser.c: removed max-min estimation method for subvolumes. We
will use the standard variance method and try to use a sufficient
number of points to estimate the variance reliably.
Wed Dec 20 21:32:40 2000 Brian Gough <bjg@network-theory.co.uk>
* vegas.c: tidied up the algorithm, deal with cases of sigma = 0
explicitly.
Sat Dec 9 14:19:53 2000 Brian Gough <bjg@network-theory.co.uk>
* reorganization and clean up
Thu Nov 16 19:50:27 2000 Brian Gough <bjg@network-theory.co.uk>
* miser.c: rewrite to fix overflows and make calling conventions
consistent
* plain.c: rewrite to fix overflows and make calling conventions
consistent
Thu Oct 26 20:06:36 2000 Brian Gough <bjg@network-theory.co.uk>
* plain.c (gsl_monte_plain_integrate): integer factor
calls*(calls-1) used in numerical expression can easily overflow,
changed to calls*(calls-1.0).
Sat Oct 21 20:36:06 2000 Brian Gough <bjg@network-theory.co.uk>
* miser.c (gsl_monte_miser_integrate): fixed bug where hits_l was
used in place of hits_r
Tue Sep 19 19:16:37 2000 Brian Gough <bjg@network-theory.co.uk>
* plain.c (gsl_monte_plain_alloc): initialise check_done to avoid
warning about use of unitialised memory
* vegas.c (gsl_monte_vegas_alloc): as above
* miser.c (gsl_monte_miser_alloc): as above
* plain.c miser.c: removed use of sprintf for error messages
Wed May 31 19:50:19 2000 Brian Gough <bjg@network-theory.co.uk>
* miser_test.c (main): increased tolerances to allow tests to pass
with different compilers
Mon May 15 15:26:22 2000 Brian Gough <bjg@network-theory.co.uk>
* vegas_test.c (main): added gsl_ieee_env_setup() to allow change
of precision in tests
* miser_test.c (main): ditto
* plain_test.c (main): ditto
Fri Feb 26 14:49:56 1999 Brian Gough <bjg@netsci.freeserve.co.uk>
* Makefile.am: removed ..._LDFLAGS = -static since this is
specific to gcc
Wed Nov 18 10:59:56 1998 Brian Gough <bjg@vvv.lanl.gov>
* use standard headers templates_on.h and templates_off.h instead
of source.h
Tue Nov 17 16:49:12 1998 Brian Gough <bjg@vvv.lanl.gov>
* added #include <config.h> to all top-level source files
* plain.c (gsl_monte_plain_integrate): replaced myMAX by GSL_MAX
* utils.c: move macros around to avoid double definition
Fri Aug 14 10:12:06 1998 Brian Gough <bjg@vvv.lanl.gov>
* Makefile.am: I needed to add utils.h to libgslmonte_a_SOURCES to
get it to work with my automake
Thu Jul 30 17:31:29 1998 booth <booth@planck.pha.jhu.edu>
* gsl_monte_miser.h, miser.c:
Turn off the annoying warning in miser unless the user requests it.
Wed Jul 29 20:24:54 1998 bjg <bjg@vvv.lanl.gov>
* Makefile.am, Makefile.in: some fixes to pass make distcheck
* Makefile.am, Makefile.in:
experimenting with new top level makefile.am
Tue Jul 28 17:05:20 1998 booth <booth@planck.pha.jhu.edu>
* plain.c, vegas.c, miser.c:
make all the _free functions check their argument. Also, the init functions
now throw EFAULT if given a null pointer.
* gsl_monte_vegas.h, vegas.c:
vegas1, vegas2 and vegas3 all go away. Get them by setting the "stage"
variabe appropriately. Also, make _free check its argument.
* vegas.c, gsl_monte_vegas.h: minor cleanup prior to be change.
* vegas.c: minor changes, commiting by accident.
Mon Jul 27 22:52:49 1998 bjg <bjg@vvv.lanl.gov>
* Makefile.in, Makefile.am:
fixed some of the include requirements for make dist
Mon Jul 27 15:19:54 1998 booth <booth@planck.pha.jhu.edu>
* vegas_print.h, vegas_test.c, miser_test.c, vegas-print.c, vegas.c, Attic/gsl_vegas.h, Attic/gsl_vegas_print.h, gsl_monte_vegas.h, miser.c, Attic/gsl_miser.h, Makefile.am, Makefile.in, gsl_monte_miser.h:
Renamed public header files to follow convention correctly.
* vegas.c, vegas_test.c, miser.c, miser_test.c, plain.c, plain_test.c, Attic/gsl_miser.h, Attic/gsl_vegas.h, gsl_monte_plain.h:
All the integration functions now end with _integrate (except for
vegas1, vegas2 and vegas3 which are going away RSN).
Tue Jul 21 21:54:33 1998 booth <booth@planck.pha.jhu.edu>
* vegas.c, vegas-print.c, Attic/gsl_vegas_print.h, Attic/gsl_vegas.h:
trivial stuff: eliminate compiler warnings, eliminate some unneeded variables,
change some types to make consistent with plain and miser.
* gsl_monte_plain.h, plain.c, plain_test.c:
Make "plain" conform to same style as miser and vegas.
Fri Jul 17 02:23:40 1998 jungman <jungman@nnn.lanl.gov>
* Makefile.in: we're gonna make it...
Thu Jul 16 16:23:45 1998 booth <booth@planck.pha.jhu.edu>
* vegas-print.c, vegas.c, Attic/gsl_vegas_print.h:
Have now completely eliminated all static variables.
* vegas_test.c, vegas.c, Attic/gsl_vegas.h, Attic/gsl_vegas_print.h, vegas-print.c:
Continuing to remove all static variables from vegas.
Wed Jul 15 19:10:24 1998 booth <booth@planck.pha.jhu.edu>
* vegas.c, vegas_test.c, Attic/gsl_vegas.h:
Do the state-structure thing for vegas. Not finished, so far the only
real content is the rng structure.
|
