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
|
Changes for 2.2.19
- Updated with a fix to write.cpp.
Changes for 2.2.18
- Updated with changes for Intel compiler v15-16.
Changes for 2.2.17
- Added an input parameter mpstrlen to mp_inp.f.
Changes for 2.2.15
- Fixed intermittent memory problem in dotd.cpp.
Changes for 2.2.6
- Fixed F-format bug in Fortran file mp_mod.f.
- Fixed mpread bug in Fortran file mp_modx.f
Changes for 2.2.2
- Clean up mp_mod.f, added mpeq_xz (assignment to complex)
and cmplx (creation of mp_complex).
- Fix dble to be more accurate, and avoid spurious underflow.
- Make C++ I/O of mp_real closer to standard C++ I/O.
- Added to_string function.
Changes for 2.2.1
- Some bug fixes for Cygwin environment.
Changes for 2.2.0
This is a major update to ARPREC, there are several API incompatibilities.
- Moved C++ main entry in libqdmod.a to libarprec_f_main.a.
This allows to link Fortran code using ARPREC with custom C++
main function. Pure Fortran code will need to be linked with
arprec_f_main library in addition to arprecmod and arprec library.
- Constructor taking integers now just initializes the data.
For example "x = mp_real(35)" now initializes x to 35, instead
of creating an mp_real with 35 words. To get the old behaviour,
do something like mp_real(0.0, 35).
- String output routines write and to_string routines now takes a single
fmtflags.
- Members are now all public, making it a plain-old-data type.
- Removed mpr_location, set/getData accessors.
- mp_real(double *) constructors are now explicit.
- Renamed several static variables in class mp to more descriptive names.
- Various constants are now in mp_real class (mp_real::_pi, _log2,
_log10, and _eps).
Changes for 2.1.104
- Fixed bug in division and increments for mp_int.
- Output correct number of digits in c_mpwrite / c_mpout.
- Fixed compilation errors and warnings.
Changes for 2.1.103
- Minor changes to accomodate Windows build.
Changes for 2.1.102
- Ignore CC compiler on Apple systems.
- Fix C++ pslq demos to print out integers when appropriate.
- Fix dependency of Fortran modules to allow parallel make.
- Fix bug in mp_real::mpnorm that was accessing past end of an array.
Changes for 2.1.101
- Added third display option in math toolkit.
- Bug fix in math toolkit.
Changes for 2.1.100
- Added fortran/Makefile.sample, a sample Makefile for using ARPREC
library with Fortran programs.
- Fixed bug in arprec-config where --fcflags had an extra @.
Changes for 2.1.99
- Added read_binary and write_binary for binary I/O.
- Fixed bug where decimal point and minus sign was not printed correctly
in certain cases.
- General improvements in I/O routines, including adjustable exponent
width and column width.
- Added mp_real constructor accepting std::string.
- Perform some I/O tests during "make check".
Changes for 2.2.3
- Removed dynamic precision from quadgs routine of Fortran and C++
versions of the Gaussian quadrature demonstration program.
- Fixed some bugs in the erf and erfc routines.
|