-
Notifications
You must be signed in to change notification settings - Fork 14
/
ChangeLog
155 lines (92 loc) · 5.01 KB
/
ChangeLog
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
ChangeLog for Simple Python Fixed-Point Module
03Jun22 - *** spfpm-1.6 released
03Jun22 - Reworked decimal-precision estimation within toDecimalString()
22May22 - Added log2() method and improved efficiency for logs far from unity
09Oct21 - Migrated from distutils to setuptools
18Sep20 - Added FXfamily method for importing two's complement numbers
29Mar20 - Extended accuracy demo plots to include Machin's formula for Pi
27Mar20 - Neatened calculation of FXfamily.exp1
Added demo plot of accuracy of exp(1)
21Mar20 - *** spfpm-1.5 released
21Mar20 - Improved testing of starting approximation within sqrt() method
15Mar20 - Improved handling of domain errors and edge-cases in __pow__() method
01Nov19 - Patched rounding behaviour on FXnum creation from float
12May18 - Added signed/twos-complement option to FXnum.toBinaryString() method
03May18 - Added 'precision' option to FXnum.toDecimalString() method
Improved estimation of number of digits in decimal representation
28Apr18 - Replaced asin() formula for pi with Bailey-Borwein-Plouffe series
16Apr18 - *** spfpm-1.4 released
13Apr18 - Added method for printing FXnum as binary/octal/hexadecimal string
11Apr18 - Added shortcut method to FXfamily for creating FXnum instances
24Feb18 - *** spfpm-1.3 released
24Jun17 - Improved convergence of asin() series.
Replaced atan() computation of pi with asin().
29May17 - Neatened asin() and acos() methods, avoiding use of atan().
28May17 - Improved efficiency of computation of log(2)
Converted README.txt into markdown
13May17 - Added optimized FXnum constructor to improve speed of arithmetic
Added command-line switches for demonstration script
Added plot of Pi accuracy to demonstration script
06May17 - *** spfpm-1.2 released
05May17 - removed need to have same FXfamily object when combining FXnums
- added '__slots__' specifier to FXnum fields
15Jul14 - *** spfpm-1.1 released
15Jul14 - added left/right shift operators
improved initialization of FXnum objects from output of repr()
14Jul14 - replaced various getter methods in FXfamily with properties
added improved expressions for computing exp(1) and log(2)
updated graph-based demos to use Matplotlib in place of Gnuplot
19May09 - *** spfpm-1.0.1 released
20Feb09 - patched FXnum constructor to check for overflows
31Jan09 - *** spfpm-1.0 released
31Jan09 - improved accuracy of pi,log(2) etc used in trig,exp functions
06Dec08 - ported to Python-3.0
11Apr08 - *** spfpm-0.6 released
09Apr08 - added validator in FXfamily to allow limited range of integer-part
07Oct07 - added basic FLOPS calculator to demo.py
27May07 - *** spfpm-0.5 released
26May07 - minor adjustments to ensure test-cases run under Python-2.2
added unit-test for repr()
21May07 - neatened module importing in unit-test routines
23Apr07 - removed use of 'isinstance' from FXfamily & (almost all of) FXnum
11Feb07 - added improved asin()/acos() methods based on 1-cos2t transformations
04Feb07 - removed (deprecated) FPnum wrappers
03Feb07 - *** spfpm-0.4 released
03Feb07 - patched rounding on cast to int/long to match behaviour of float-casts
added unit-test for sqrt() method
expanded set of exception classes, including domain-error etc
added unit-test for asin() & acos() methods
02Feb07 - added demo/test/clean Makefile targets
01Feb07 - added setup.py distutils-script
31Jan07 - added basic asin() & acos() methods
30Jan07 - replaced _rawsincos() method with dual-mode _rawQsine() method
added sin() & cos() methods
29Jan07 - patched spurious family-conversion in __neg__() operator
27Jan07 - extended discusions + examples in documentation
added doctest.testmod() test for documentation
added __abs__, __pos__, __hash__() methods
changed default distribution-format to .zip in place of .tar.gz
22Jan07 - neatened internal usage of truncating division
wired-together __div__ & __truediv__ operators
21Jan07 - *** spfpm-0.3 released
21Jan06 - added backwards-compatibility SetFraction method to FPnum wrapper
renamed arctan() method to atan() to agree with math-module convention
20Jan06 - added unit-test for resolution-conversions
added arctan & _rawarctan() methods
adjusted sincos() method to reduce angles moduli Pi rather than halving
added GetPi() method to FXfamily
adjusted FXfamily to use greater bit-count when finding cached constants
19Jan06 - changed 'FPnum' to 'FXnum' to avoid abbreviation-clash with 'Floating Point'
added FPnum backward-compatibility wrapper
18Jan06 - added unit-test for __str__ method
17Jan06 - converted resolution information to shareable FPfamily object
16Jan06 - added test for __nonzero__ method & patched reference to 'other'
removed __iadd__, __isub__ methods to make immutable object
added unit-test for immutability
15Jan07 - *** spfpm-0.2 released
13Jan07 - added basic Makefile for building release packages
added unit-tests for pow() and sincos() methods
08Jan07 - added logarithm methods
03Jan07 - added basic unit-testing class for arithmetic operations
02Jan07 - converted RCS repository into SVN
26Dec06 - *** prototype of spfpm-0.1 released