line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
############################################################################# |
2
|
|
|
|
|
|
|
# Math/Big.pm -- usefull routines with Big numbers (BigInt/BigFloat) |
3
|
|
|
|
|
|
|
|
4
|
|
|
|
|
|
|
package Math::Big; |
5
|
|
|
|
|
|
|
|
6
|
|
|
|
|
|
|
require 5.006002; # anything lower is simple untested |
7
|
|
|
|
|
|
|
|
8
|
3
|
|
|
3
|
|
142749
|
use strict; |
|
3
|
|
|
|
|
31
|
|
|
3
|
|
|
|
|
98
|
|
9
|
3
|
|
|
3
|
|
17
|
use warnings; |
|
3
|
|
|
|
|
5
|
|
|
3
|
|
|
|
|
105
|
|
10
|
|
|
|
|
|
|
|
11
|
3
|
|
|
3
|
|
2610
|
use Math::BigInt '1.97'; |
|
3
|
|
|
|
|
62729
|
|
|
3
|
|
|
|
|
14
|
|
12
|
3
|
|
|
3
|
|
55346
|
use Math::BigFloat; |
|
3
|
|
|
|
|
62163
|
|
|
3
|
|
|
|
|
17
|
|
13
|
3
|
|
|
3
|
|
854
|
use Exporter; |
|
3
|
|
|
|
|
7
|
|
|
3
|
|
|
|
|
9842
|
|
14
|
|
|
|
|
|
|
|
15
|
|
|
|
|
|
|
our $VERSION = '1.15'; |
16
|
|
|
|
|
|
|
our @ISA = qw( Exporter ); |
17
|
|
|
|
|
|
|
our @EXPORT_OK = qw( primes fibonacci base to_base hailstone factorial |
18
|
|
|
|
|
|
|
euler bernoulli pi log |
19
|
|
|
|
|
|
|
tan cos sin cosh sinh arctan arctanh arcsin arcsinh |
20
|
|
|
|
|
|
|
); |
21
|
|
|
|
|
|
|
|
22
|
|
|
|
|
|
|
# some often used constants: |
23
|
|
|
|
|
|
|
my $four = Math::BigFloat->new(4); |
24
|
|
|
|
|
|
|
my $sixteen = Math::BigFloat->new(16); |
25
|
|
|
|
|
|
|
my $fone = Math::BigFloat->bone(); # pi |
26
|
|
|
|
|
|
|
my $one = Math::BigInt->bone(); # hailstone, sin, cos etc |
27
|
|
|
|
|
|
|
my $two = Math::BigInt->new(2); # hailstone, sin, cos etc |
28
|
|
|
|
|
|
|
my $three = Math::BigInt->new(3); # hailstone |
29
|
|
|
|
|
|
|
|
30
|
|
|
|
|
|
|
my $five = Math::BigFloat->new(5); # for pi |
31
|
|
|
|
|
|
|
my $twothreenine = Math::BigFloat->new(239); # for pi |
32
|
|
|
|
|
|
|
|
33
|
|
|
|
|
|
|
# In scalar context this returns the prime count (# of primes <= N). |
34
|
|
|
|
|
|
|
# In array context it returns a list of primes from 2 to N. |
35
|
|
|
|
|
|
|
sub primes { |
36
|
16
|
|
|
16
|
1
|
5049
|
my $end = shift; |
37
|
16
|
50
|
|
|
|
45
|
return unless defined $end; |
38
|
16
|
50
|
|
|
|
42
|
$end = $end->numify() if ref($end) =~ /^Math::Big/; |
39
|
16
|
0
|
|
|
|
41
|
if ($end < 2) { return !wantarray ? Math::BigInt->bzero() : (); } |
|
0
|
50
|
|
|
|
0
|
|
40
|
16
|
0
|
|
|
|
37
|
if ($end < 3) { return !wantarray ? $one->copy : ($two->copy); } |
|
0
|
50
|
|
|
|
0
|
|
41
|
16
|
100
|
|
|
|
30
|
if ($end < 5) { return !wantarray ? $two->copy : ($two->copy, $three->copy); } |
|
3
|
100
|
|
|
|
19
|
|
42
|
|
|
|
|
|
|
|
43
|
13
|
100
|
|
|
|
38
|
$end-- unless ($end & 1); |
44
|
13
|
|
|
|
|
25
|
my $s_end = $end >> 1; |
45
|
13
|
|
|
|
|
38
|
my $whole = int( ($end>>1) / 15); |
46
|
|
|
|
|
|
|
# Be conservative. This would result in terabytes of array output. |
47
|
13
|
50
|
|
|
|
31
|
die "Cannot return $end primes!" if $whole > 1_145_324_612; # ~32 GB string |
48
|
13
|
|
|
|
|
37
|
my $sieve = "100010010010110" . "011010010010110" x $whole; |
49
|
13
|
|
|
|
|
27
|
substr($sieve, $s_end+1) = ''; # Clip to the right number of entries |
50
|
13
|
|
|
|
|
37
|
my ($n, $limit) = ( 7, int(sqrt($end)) ); |
51
|
13
|
|
|
|
|
32
|
while ( $n <= $limit ) { |
52
|
0
|
|
|
|
|
0
|
for (my $s = ($n*$n) >> 1; $s <= $s_end; $s += $n) { |
53
|
0
|
|
|
|
|
0
|
substr($sieve, $s, 1) = '1'; |
54
|
|
|
|
|
|
|
} |
55
|
0
|
|
|
|
|
0
|
do { $n += 2 } while substr($sieve, $n>>1, 1); |
|
0
|
|
|
|
|
0
|
|
56
|
|
|
|
|
|
|
} |
57
|
|
|
|
|
|
|
|
58
|
13
|
100
|
|
|
|
31
|
return Math::BigInt->new(1 + $sieve =~ tr/0//) if !wantarray; |
59
|
|
|
|
|
|
|
|
60
|
12
|
|
|
|
|
26
|
my @primes = (2, 3, 5); |
61
|
12
|
|
|
|
|
20
|
$n = 7-2; |
62
|
12
|
|
|
|
|
48
|
foreach my $s (split("0", substr($sieve, 3), -1)) { |
63
|
44
|
|
|
|
|
64
|
$n += 2 + 2 * length($s); |
64
|
44
|
100
|
|
|
|
88
|
push @primes, $n if $n <= $end; |
65
|
|
|
|
|
|
|
} |
66
|
12
|
|
|
|
|
27
|
return map { Math::BigInt->new($_) } @primes; |
|
69
|
|
|
|
|
2052
|
|
67
|
|
|
|
|
|
|
} |
68
|
|
|
|
|
|
|
|
69
|
|
|
|
|
|
|
sub fibonacci |
70
|
|
|
|
|
|
|
{ |
71
|
22
|
|
|
22
|
1
|
25226
|
my $x = shift; |
72
|
22
|
50
|
33
|
|
|
107
|
$x = Math::BigInt -> new($x) |
73
|
|
|
|
|
|
|
unless ref($x) && $x -> isa("Math::BigInt"); |
74
|
22
|
|
|
|
|
1073
|
$x -> bfib(); |
75
|
|
|
|
|
|
|
} |
76
|
|
|
|
|
|
|
|
77
|
|
|
|
|
|
|
sub base |
78
|
|
|
|
|
|
|
{ |
79
|
7
|
|
|
7
|
1
|
4819
|
my ($number,$base) = @_; |
80
|
|
|
|
|
|
|
|
81
|
7
|
50
|
|
|
|
37
|
$number = Math::BigInt->new($number) unless ref $number; |
82
|
7
|
50
|
|
|
|
314
|
$base = Math::BigInt->new($base) unless ref $base; |
83
|
|
|
|
|
|
|
|
84
|
7
|
50
|
|
|
|
249
|
return if $number < $base; |
85
|
7
|
|
|
|
|
260
|
my $n = Math::BigInt->new(0); |
86
|
7
|
|
|
|
|
751
|
my $trial = $base; |
87
|
|
|
|
|
|
|
# 9 = 2**3 + 1 |
88
|
7
|
|
|
|
|
17
|
while ($trial < $number) |
89
|
|
|
|
|
|
|
{ |
90
|
20
|
|
|
|
|
1057
|
$trial *= $base; $n++; |
|
20
|
|
|
|
|
1350
|
|
91
|
|
|
|
|
|
|
} |
92
|
7
|
|
|
|
|
472
|
$trial /= $base; $a = $number - $trial; |
|
7
|
|
|
|
|
1060
|
|
93
|
7
|
|
|
|
|
921
|
($n,$a); |
94
|
|
|
|
|
|
|
} |
95
|
|
|
|
|
|
|
|
96
|
|
|
|
|
|
|
sub to_base |
97
|
|
|
|
|
|
|
{ |
98
|
10
|
|
|
10
|
1
|
10232
|
my $x = shift; |
99
|
10
|
50
|
33
|
|
|
54
|
$x = Math::BigInt->new($x) |
100
|
|
|
|
|
|
|
unless ref($x) && $x -> isa("Math::BigInt"); |
101
|
10
|
|
|
|
|
589
|
$x -> to_base(@_); |
102
|
|
|
|
|
|
|
} |
103
|
|
|
|
|
|
|
|
104
|
|
|
|
|
|
|
sub hailstone |
105
|
|
|
|
|
|
|
{ |
106
|
|
|
|
|
|
|
# return in list context the hailstone sequence, in scalar context the |
107
|
|
|
|
|
|
|
# number of steps to reach 1 |
108
|
11
|
|
|
11
|
1
|
12195
|
my ($n) = @_; |
109
|
|
|
|
|
|
|
|
110
|
11
|
50
|
|
|
|
54
|
$n = Math::BigInt->new($n) unless ref $n; |
111
|
|
|
|
|
|
|
|
112
|
11
|
50
|
33
|
|
|
485
|
return if $n->is_nan() || $n->is_negative(); |
113
|
|
|
|
|
|
|
|
114
|
|
|
|
|
|
|
# Use the Math::BigInt lib directly for more speed, since all numbers |
115
|
|
|
|
|
|
|
# involved are positive integers. |
116
|
|
|
|
|
|
|
|
117
|
11
|
|
|
|
|
175
|
my $lib = Math::BigInt->config()->{lib}; |
118
|
11
|
|
|
|
|
499
|
$n = $n->{value}; |
119
|
11
|
|
|
|
|
24
|
my $three_ = $three->{value}; |
120
|
11
|
|
|
|
|
16
|
my $two_ = $two->{value}; |
121
|
|
|
|
|
|
|
|
122
|
11
|
100
|
|
|
|
23
|
if (wantarray) |
123
|
|
|
|
|
|
|
{ |
124
|
4
|
|
|
|
|
6
|
my @seq; |
125
|
4
|
|
|
|
|
11
|
while (! $lib->_is_one($n)) |
126
|
|
|
|
|
|
|
{ |
127
|
|
|
|
|
|
|
# push @seq, Math::BigInt->new( $lib->_str($n) ); |
128
|
47
|
|
|
|
|
563
|
push @seq, bless { value => $lib->_copy($n), sign => '+' }, "Math::BigInt"; |
129
|
|
|
|
|
|
|
|
130
|
|
|
|
|
|
|
# was: ($n->is_odd()) ? ($n = $n * 3 + 1) : ($n = $n / 2); |
131
|
47
|
100
|
|
|
|
331
|
if ($lib->_is_odd($n)) |
132
|
|
|
|
|
|
|
{ |
133
|
19
|
|
|
|
|
78
|
$n = $lib->_mul ($n, $three_); $n = $lib->_inc ($n); |
|
19
|
|
|
|
|
165
|
|
134
|
|
|
|
|
|
|
|
135
|
|
|
|
|
|
|
# We now know that $n is at least 10 ( (3 * 3) + 1 ) because $n > 1 |
136
|
|
|
|
|
|
|
# before we entered, and since $n was odd, it must have been at least |
137
|
|
|
|
|
|
|
# 3. So the next step is $n /= 2: |
138
|
19
|
|
|
|
|
179
|
push @seq, bless { value => $lib->_copy($n), sign => '+' }, "Math::BigInt"; |
139
|
|
|
|
|
|
|
# this is better, but slower: |
140
|
|
|
|
|
|
|
#push @seq, Math::BigInt->new( $lib->_str($n) ); |
141
|
|
|
|
|
|
|
# next step is $n /= 2 as usual (we save the else {} block, too) |
142
|
|
|
|
|
|
|
} |
143
|
47
|
|
|
|
|
239
|
$n = $lib->_div($n, $two_); |
144
|
|
|
|
|
|
|
} |
145
|
4
|
|
|
|
|
52
|
push @seq, Math::BigInt->bone(); |
146
|
4
|
|
|
|
|
275
|
return @seq; |
147
|
|
|
|
|
|
|
} |
148
|
|
|
|
|
|
|
|
149
|
7
|
|
|
|
|
9
|
my $i = 1; |
150
|
7
|
|
|
|
|
21
|
while (! $lib->_is_one($n)) |
151
|
|
|
|
|
|
|
{ |
152
|
47
|
|
|
|
|
497
|
$i++; |
153
|
|
|
|
|
|
|
# was: ($n->is_odd()) ? ($n = $n * 3 + 1) : ($n = $n / 2); |
154
|
47
|
100
|
|
|
|
74
|
if ($lib->_is_odd($n)) |
155
|
|
|
|
|
|
|
{ |
156
|
18
|
|
|
|
|
68
|
$n = $lib->_mul ($n, $three_); $n = $lib->_inc ($n); |
|
18
|
|
|
|
|
152
|
|
157
|
|
|
|
|
|
|
|
158
|
|
|
|
|
|
|
# We now know that $n is at least 10 ( (3 * 3) + 1 ) because $n > 1 |
159
|
|
|
|
|
|
|
# before we entered, and since $n was odd, it must have been at least 3. |
160
|
|
|
|
|
|
|
# So the next step is $n /= 2 as usual (we save the else {} block, too). |
161
|
18
|
|
|
|
|
105
|
$i++; # one more (we know that $n cannot be 1) |
162
|
|
|
|
|
|
|
} |
163
|
47
|
|
|
|
|
142
|
$n = $lib->_div($n, $two_); |
164
|
|
|
|
|
|
|
} |
165
|
7
|
|
|
|
|
90
|
Math::BigInt->new($i); |
166
|
|
|
|
|
|
|
} |
167
|
|
|
|
|
|
|
|
168
|
|
|
|
|
|
|
sub factorial |
169
|
|
|
|
|
|
|
{ |
170
|
|
|
|
|
|
|
# calculate n! - use Math::BigInt bfac() for speed |
171
|
7
|
|
|
7
|
1
|
5426
|
my ($n) = shift; |
172
|
|
|
|
|
|
|
|
173
|
7
|
50
|
|
|
|
18
|
if (ref($n)) |
174
|
|
|
|
|
|
|
{ |
175
|
0
|
|
|
|
|
0
|
$n->copy()->bfac(); |
176
|
|
|
|
|
|
|
} |
177
|
|
|
|
|
|
|
else |
178
|
|
|
|
|
|
|
{ |
179
|
7
|
|
|
|
|
23
|
Math::BigInt->new($n)->bfac(); |
180
|
|
|
|
|
|
|
} |
181
|
|
|
|
|
|
|
} |
182
|
|
|
|
|
|
|
|
183
|
|
|
|
|
|
|
sub bernoulli |
184
|
|
|
|
|
|
|
{ |
185
|
|
|
|
|
|
|
# returns the nth Bernoulli number. In scalar context as Math::BigFloat |
186
|
|
|
|
|
|
|
# fraction, in list context as two Math:BigFloats, which, if divided, give |
187
|
|
|
|
|
|
|
# the same result. The series runs this: |
188
|
|
|
|
|
|
|
# 1/6, 1/30, 1/42, 1/30, 5/66, 691/2730, etc |
189
|
|
|
|
|
|
|
|
190
|
|
|
|
|
|
|
# Since I do not have yet a way to compute this, I have a table of the |
191
|
|
|
|
|
|
|
# first 40. So bernoulli(41) will fail for now. |
192
|
|
|
|
|
|
|
|
193
|
61
|
|
|
61
|
1
|
29942
|
my $n = shift; |
194
|
|
|
|
|
|
|
|
195
|
61
|
50
|
|
|
|
490
|
return if $n < 0; |
196
|
61
|
|
|
|
|
118
|
my @table_1 = ( 1,1, -1,2 ); # 0, 1 |
197
|
61
|
|
|
|
|
217
|
my @table = ( |
198
|
|
|
|
|
|
|
1,6, -1,30, 1,42, -1,30, 5,66, -691,2730, # 2, 4, |
199
|
|
|
|
|
|
|
7,6, -3617,510, 43867,798, |
200
|
|
|
|
|
|
|
-174611,330, |
201
|
|
|
|
|
|
|
854513,138, |
202
|
|
|
|
|
|
|
'-236364091',2730, |
203
|
|
|
|
|
|
|
'8553103',6, |
204
|
|
|
|
|
|
|
'-23749461029',870, |
205
|
|
|
|
|
|
|
'8615841276005',14322, |
206
|
|
|
|
|
|
|
'-7709321041217',510, |
207
|
|
|
|
|
|
|
'2577687858367',6, |
208
|
|
|
|
|
|
|
'-26315271553053477373',1919190, |
209
|
|
|
|
|
|
|
'2929993913841559',6, |
210
|
|
|
|
|
|
|
'-261082718496449122051',13530, # 40 |
211
|
|
|
|
|
|
|
); |
212
|
61
|
|
|
|
|
79
|
my ($a,$b); |
213
|
61
|
100
|
|
|
|
170
|
if ($n < 2) |
|
|
100
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
214
|
|
|
|
|
|
|
{ |
215
|
2
|
|
|
|
|
11
|
$a = Math::BigFloat->new($table_1[$n*2]); |
216
|
2
|
|
|
|
|
231
|
$b = Math::BigFloat->new($table_1[$n*2+1]); |
217
|
|
|
|
|
|
|
} |
218
|
|
|
|
|
|
|
# n is odd: |
219
|
|
|
|
|
|
|
elsif (($n & 1) == 1) |
220
|
|
|
|
|
|
|
{ |
221
|
20
|
|
|
|
|
63
|
$a = Math::BigFloat->bzero(); |
222
|
20
|
|
|
|
|
1102
|
$b = Math::BigFloat->bone(); |
223
|
|
|
|
|
|
|
} |
224
|
|
|
|
|
|
|
elsif ($n <= 40) |
225
|
|
|
|
|
|
|
{ |
226
|
39
|
|
|
|
|
51
|
$n -= 2; |
227
|
39
|
|
|
|
|
108
|
$a = Math::BigFloat->new($table[$n]); |
228
|
39
|
|
|
|
|
3713
|
$b = Math::BigFloat->new($table[$n+1]); |
229
|
|
|
|
|
|
|
} |
230
|
|
|
|
|
|
|
else |
231
|
|
|
|
|
|
|
{ |
232
|
0
|
0
|
|
|
|
0
|
die 'Bernoulli numbers over 40 not yet implemented.' if $n > 40; |
233
|
|
|
|
|
|
|
} |
234
|
61
|
50
|
|
|
|
7414
|
wantarray ? ($a,$b): $a/$b; |
235
|
|
|
|
|
|
|
} |
236
|
|
|
|
|
|
|
|
237
|
|
|
|
|
|
|
sub euler |
238
|
|
|
|
|
|
|
{ |
239
|
|
|
|
|
|
|
# Calculate Euler's number. |
240
|
|
|
|
|
|
|
# first argument is x, so that result is e ** x |
241
|
|
|
|
|
|
|
# Second argument is accuracy (number of significant digits), it |
242
|
|
|
|
|
|
|
# stops when at least so much plus one digits are 'stable' and then |
243
|
|
|
|
|
|
|
# rounds it. Default is 42. |
244
|
4
|
|
|
4
|
1
|
6307
|
my $x = $_[0]; |
245
|
4
|
50
|
33
|
|
|
42
|
$x = Math::BigFloat->new($x) if !ref($x) || (!$x->isa('Math::BigFloat')); |
246
|
|
|
|
|
|
|
|
247
|
4
|
|
|
|
|
565
|
$x->bexp($_[1]); |
248
|
|
|
|
|
|
|
} |
249
|
|
|
|
|
|
|
|
250
|
|
|
|
|
|
|
sub sin |
251
|
|
|
|
|
|
|
{ |
252
|
|
|
|
|
|
|
# calculate sinus |
253
|
|
|
|
|
|
|
# first argument is x, so that result is sin(x) |
254
|
|
|
|
|
|
|
# Second argument is accuracy (number of significant digits), it |
255
|
|
|
|
|
|
|
# stops when at least so much plus one digits are 'stable' and then |
256
|
|
|
|
|
|
|
# rounds it. Default is 42. |
257
|
8
|
50
|
|
8
|
1
|
8046
|
my $x = shift; $x = 0 if !defined $x; |
|
8
|
|
|
|
|
25
|
|
258
|
8
|
|
50
|
|
|
25
|
my $d = abs(shift || 42); $d = abs($d)+1; |
|
8
|
|
|
|
|
11
|
|
259
|
|
|
|
|
|
|
|
260
|
8
|
50
|
|
|
|
38
|
$x = Math::BigFloat->new($x) if ref($x) ne 'Math::BigFloat'; |
261
|
|
|
|
|
|
|
|
262
|
|
|
|
|
|
|
# taylor: x^3 x^5 x^7 x^9 |
263
|
|
|
|
|
|
|
# sin = x - --- + --- - --- + --- ... |
264
|
|
|
|
|
|
|
# 3! 5! 7! 9! |
265
|
|
|
|
|
|
|
|
266
|
|
|
|
|
|
|
# difference for each term is thus x^2 and 1,2 |
267
|
|
|
|
|
|
|
|
268
|
8
|
|
|
|
|
1283
|
my $sin = $x->copy(); my $last = 0; |
|
8
|
|
|
|
|
221
|
|
269
|
8
|
|
|
|
|
11
|
my $sign = 1; # start with -= |
270
|
8
|
|
|
|
|
23
|
my $x2 = $x * $x; # X ^ 2, difference between terms |
271
|
8
|
|
|
|
|
1059
|
my $over = $x2 * $x; # X ^ 3 |
272
|
8
|
|
|
|
|
972
|
my $below = Math::BigFloat->new(6); my $factorial = Math::BigFloat->new(4); |
|
8
|
|
|
|
|
815
|
|
273
|
8
|
|
|
|
|
758
|
while ($sin->bcmp($last) != 0) # no $x-$last > $diff because bdiv() limit on accuracy |
274
|
|
|
|
|
|
|
{ |
275
|
46
|
|
|
|
|
11188
|
$last = $sin->copy(); |
276
|
46
|
100
|
|
|
|
1371
|
if ($sign == 0) |
277
|
|
|
|
|
|
|
{ |
278
|
22
|
|
|
|
|
45
|
$sin += $over->copy()->bdiv($below,$d); |
279
|
|
|
|
|
|
|
} |
280
|
|
|
|
|
|
|
else |
281
|
|
|
|
|
|
|
{ |
282
|
24
|
|
|
|
|
47
|
$sin -= $over->copy()->bdiv($below,$d); |
283
|
|
|
|
|
|
|
} |
284
|
46
|
|
|
|
|
62693
|
$sign = 1-$sign; # alternate |
285
|
46
|
|
|
|
|
118
|
$over *= $x2; # $x*$x |
286
|
46
|
|
|
|
|
4867
|
$below *= $factorial; $factorial++; # n*(n+1) |
|
46
|
|
|
|
|
4544
|
|
287
|
46
|
|
|
|
|
4106
|
$below *= $factorial; $factorial++; |
|
46
|
|
|
|
|
5551
|
|
288
|
|
|
|
|
|
|
} |
289
|
8
|
|
|
|
|
1875
|
$sin->bround($d-1); |
290
|
|
|
|
|
|
|
} |
291
|
|
|
|
|
|
|
|
292
|
|
|
|
|
|
|
sub cos |
293
|
|
|
|
|
|
|
{ |
294
|
|
|
|
|
|
|
# calculate cosinus |
295
|
|
|
|
|
|
|
# first argument is x, so that result is cos(x) |
296
|
|
|
|
|
|
|
# Second argument is accuracy (number of significant digits), it |
297
|
|
|
|
|
|
|
# stops when at least so much plus one digits are 'stable' and then |
298
|
|
|
|
|
|
|
# rounds it. Default is 42. |
299
|
8
|
50
|
|
8
|
1
|
8283
|
my $x = shift; $x = 0 if !defined $x; |
|
8
|
|
|
|
|
46
|
|
300
|
8
|
|
50
|
|
|
25
|
my $d = abs(shift || 42); $d = abs($d)+1; |
|
8
|
|
|
|
|
11
|
|
301
|
|
|
|
|
|
|
|
302
|
8
|
50
|
|
|
|
36
|
$x = Math::BigFloat->new($x) if ref($x) ne 'Math::BigFloat'; |
303
|
|
|
|
|
|
|
|
304
|
|
|
|
|
|
|
# taylor: x^2 x^4 x^6 x^8 |
305
|
|
|
|
|
|
|
# cos = 1 - --- + --- - --- + --- ... |
306
|
|
|
|
|
|
|
# 2! 4! 6! 8! |
307
|
|
|
|
|
|
|
|
308
|
|
|
|
|
|
|
# difference for each term is thus x^2 and 1,2 |
309
|
|
|
|
|
|
|
|
310
|
8
|
|
|
|
|
1304
|
my $cos = Math::BigFloat->bone(); my $last = 0; |
|
8
|
|
|
|
|
458
|
|
311
|
8
|
|
|
|
|
23
|
my $over = $x * $x; # X ^ 2 |
312
|
8
|
|
|
|
|
1077
|
my $x2 = $over->copy(); # X ^ 2; difference between terms |
313
|
8
|
|
|
|
|
223
|
my $sign = 1; # start with -= |
314
|
8
|
|
|
|
|
22
|
my $below = Math::BigFloat->new(2); my $factorial = Math::BigFloat->new(3); |
|
8
|
|
|
|
|
821
|
|
315
|
8
|
|
|
|
|
757
|
while ($cos->bcmp($last) != 0) # no $x-$last > $diff because bdiv() limit on accuracy |
316
|
|
|
|
|
|
|
{ |
317
|
52
|
|
|
|
|
13204
|
$last = $cos->copy(); |
318
|
52
|
100
|
|
|
|
1508
|
if ($sign == 0) |
319
|
|
|
|
|
|
|
{ |
320
|
24
|
|
|
|
|
44
|
$cos += $over->copy()->bdiv($below,$d); |
321
|
|
|
|
|
|
|
} |
322
|
|
|
|
|
|
|
else |
323
|
|
|
|
|
|
|
{ |
324
|
28
|
|
|
|
|
55
|
$cos -= $over->copy()->bdiv($below,$d); |
325
|
|
|
|
|
|
|
} |
326
|
52
|
|
|
|
|
66288
|
$sign = 1-$sign; # alternate |
327
|
52
|
|
|
|
|
127
|
$over *= $x2; # $x*$x |
328
|
52
|
|
|
|
|
5757
|
$below *= $factorial; $factorial++; # n*(n+1) |
|
52
|
|
|
|
|
5971
|
|
329
|
52
|
|
|
|
|
4635
|
$below *= $factorial; $factorial++; |
|
52
|
|
|
|
|
5117
|
|
330
|
|
|
|
|
|
|
} |
331
|
8
|
|
|
|
|
1734
|
$cos->round($d-1); |
332
|
|
|
|
|
|
|
} |
333
|
|
|
|
|
|
|
|
334
|
|
|
|
|
|
|
sub tan |
335
|
|
|
|
|
|
|
{ |
336
|
|
|
|
|
|
|
# calculate tangens |
337
|
|
|
|
|
|
|
# first argument is x, so that result is tan(x) |
338
|
|
|
|
|
|
|
# Second argument is accuracy (number of significant digits), it |
339
|
|
|
|
|
|
|
# stops when at least so much plus one digits are 'stable' and then |
340
|
|
|
|
|
|
|
# rounds it. Default is 42. |
341
|
3
|
50
|
|
3
|
1
|
3948
|
my $x = shift; $x = 0 if !defined $x; |
|
3
|
|
|
|
|
9
|
|
342
|
3
|
|
50
|
|
|
11
|
my $d = abs(shift || 42); $d = abs($d)+1; |
|
3
|
|
|
|
|
6
|
|
343
|
|
|
|
|
|
|
|
344
|
3
|
50
|
|
|
|
15
|
$x = Math::BigFloat->new($x) if ref($x) ne 'Math::BigFloat'; |
345
|
|
|
|
|
|
|
|
346
|
|
|
|
|
|
|
# taylor: 1 2 3 4 5 |
347
|
|
|
|
|
|
|
|
348
|
|
|
|
|
|
|
# x^3 x^5 x^7 x^9 |
349
|
|
|
|
|
|
|
# tan = x + 1 * ----- + 2 * ----- + 17 * ----- + 62 * ----- ... |
350
|
|
|
|
|
|
|
# 3 15 315 2835 |
351
|
|
|
|
|
|
|
# |
352
|
|
|
|
|
|
|
# 2^2n * ( 2^2n - 1) * Bn * x^(2n-1) 256*255 * 1 * x^7 17 |
353
|
|
|
|
|
|
|
# ---------------------------------- : n=4: ----------------- = --- * x^7 |
354
|
|
|
|
|
|
|
# (2n)! 40320 * 30 315 |
355
|
|
|
|
|
|
|
# |
356
|
|
|
|
|
|
|
# 8! = 40320, B4 (Bernoully number 4) = 1/30 |
357
|
|
|
|
|
|
|
|
358
|
|
|
|
|
|
|
# for each term we need: 2^2n, but if we have 2^2(n-1) we use n = (n-1)*2 |
359
|
|
|
|
|
|
|
# 2 copy, 7 bmul, 2 bdiv, 3 badd, 1 bernoulli |
360
|
|
|
|
|
|
|
|
361
|
3
|
|
|
|
|
371
|
my $tan = $x->copy(); my $last = 0; |
|
3
|
|
|
|
|
84
|
|
362
|
3
|
|
|
|
|
12
|
my $x2 = $x*$x; |
363
|
3
|
|
|
|
|
394
|
my $over = $x2*$x; |
364
|
3
|
|
|
|
|
337
|
my $below = Math::BigFloat->new(24); # (1*2*3*4) (2n)! |
365
|
3
|
|
|
|
|
228
|
my $factorial = Math::BigFloat->new(5); # for next (2n)! |
366
|
3
|
|
|
|
|
304
|
my $two_n = Math::BigFloat->new(16); # 2^2n |
367
|
3
|
|
|
|
|
196
|
my $two_factor = Math::BigFloat->new(4); # 2^2(n+1) = $two_n * $two_factor |
368
|
3
|
|
|
|
|
286
|
my ($b,$b1,$b2); $b = 4; |
|
3
|
|
|
|
|
4
|
|
369
|
3
|
|
|
|
|
9
|
while ($tan->bcmp($last) != 0) # no $x-$last > $diff because bdiv() limit on accuracy |
370
|
|
|
|
|
|
|
{ |
371
|
19
|
|
|
|
|
3050
|
$last = $tan->copy(); |
372
|
19
|
|
|
|
|
578
|
($b1,$b2) = bernoulli($b); |
373
|
19
|
|
|
|
|
88
|
$tan += $over->copy()->bmul($two_n)->bmul($two_n - $fone)->bmul($b1->babs())->bdiv($below,$d)->bdiv($b2,$d); |
374
|
19
|
|
|
|
|
51327
|
$over *= $x2; # x^3, x^5 etc |
375
|
19
|
|
|
|
|
1919
|
$below *= $factorial; $factorial++; # n*(n+1) |
|
19
|
|
|
|
|
2774
|
|
376
|
19
|
|
|
|
|
1764
|
$below *= $factorial; $factorial++; |
|
19
|
|
|
|
|
2192
|
|
377
|
19
|
|
|
|
|
1861
|
$two_n *= $two_factor; # 2^2(n+1) = 2^2n * 4 |
378
|
19
|
|
|
|
|
1891
|
$b += 2; # next bernoulli index |
379
|
19
|
100
|
|
|
|
71
|
last if $b > 40; # safeguard |
380
|
|
|
|
|
|
|
} |
381
|
3
|
|
|
|
|
431
|
$tan->round($d-1); |
382
|
|
|
|
|
|
|
} |
383
|
|
|
|
|
|
|
|
384
|
|
|
|
|
|
|
sub sinh |
385
|
|
|
|
|
|
|
{ |
386
|
|
|
|
|
|
|
# calculate sinus hyperbolicus |
387
|
|
|
|
|
|
|
# first argument is x, so that result is sinh(x) |
388
|
|
|
|
|
|
|
# Second argument is accuracy (number of significant digits), it |
389
|
|
|
|
|
|
|
# stops when at least so much plus one digits are 'stable' and then |
390
|
|
|
|
|
|
|
# rounds it. Default is 42. |
391
|
2
|
50
|
|
2
|
1
|
1470
|
my $x = shift; $x = 0 if !defined $x; |
|
2
|
|
|
|
|
7
|
|
392
|
2
|
|
50
|
|
|
8
|
my $d = abs(shift || 42); $d = abs($d)+1; |
|
2
|
|
|
|
|
3
|
|
393
|
|
|
|
|
|
|
|
394
|
2
|
50
|
|
|
|
11
|
$x = Math::BigFloat->new($x) if ref($x) ne 'Math::BigFloat'; |
395
|
|
|
|
|
|
|
|
396
|
|
|
|
|
|
|
# taylor: x^3 x^5 x^7 |
397
|
|
|
|
|
|
|
# sinh = x + --- + --- + --- ... |
398
|
|
|
|
|
|
|
# 3! 5! 7! |
399
|
|
|
|
|
|
|
|
400
|
|
|
|
|
|
|
# difference for each term is thus x^2 and 1,2 |
401
|
|
|
|
|
|
|
|
402
|
2
|
|
|
|
|
236
|
my $sinh = $x->copy(); my $last = 0; |
|
2
|
|
|
|
|
83
|
|
403
|
2
|
|
|
|
|
9
|
my $x2 = $x*$x; |
404
|
2
|
|
|
|
|
273
|
my $over = $x2 * $x; my $below = Math::BigFloat->new(6); my $factorial = Math::BigFloat->new(4); |
|
2
|
|
|
|
|
224
|
|
|
2
|
|
|
|
|
204
|
|
405
|
2
|
|
|
|
|
197
|
while ($sinh->bcmp($last)) # no $x-$last > $diff because bdiv() limit on accuracy |
406
|
|
|
|
|
|
|
{ |
407
|
0
|
|
|
|
|
0
|
$last = $sinh->copy(); |
408
|
0
|
|
|
|
|
0
|
$sinh += $over->copy()->bdiv($below,$d); |
409
|
0
|
|
|
|
|
0
|
$over *= $x2; # $x*$x |
410
|
0
|
|
|
|
|
0
|
$below *= $factorial; $factorial++; # n*(n+1) |
|
0
|
|
|
|
|
0
|
|
411
|
0
|
|
|
|
|
0
|
$below *= $factorial; $factorial++; |
|
0
|
|
|
|
|
0
|
|
412
|
|
|
|
|
|
|
} |
413
|
2
|
|
|
|
|
417
|
$sinh->bround($d-1); |
414
|
|
|
|
|
|
|
} |
415
|
|
|
|
|
|
|
|
416
|
|
|
|
|
|
|
sub cosh |
417
|
|
|
|
|
|
|
{ |
418
|
|
|
|
|
|
|
# calculate cosinus hyperbolicus |
419
|
|
|
|
|
|
|
# first argument is x, so that result is cosh(x) |
420
|
|
|
|
|
|
|
# Second argument is accuracy (number of significant digits), it |
421
|
|
|
|
|
|
|
# stops when at least so much plus one digits are 'stable' and then |
422
|
|
|
|
|
|
|
# rounds it. Default is 42. |
423
|
0
|
0
|
|
0
|
1
|
0
|
my $x = shift; $x = 0 if !defined $x; |
|
0
|
|
|
|
|
0
|
|
424
|
0
|
|
0
|
|
|
0
|
my $d = abs(shift || 42); $d = abs($d)+1; |
|
0
|
|
|
|
|
0
|
|
425
|
|
|
|
|
|
|
|
426
|
0
|
0
|
|
|
|
0
|
$x = Math::BigFloat->new($x) if ref($x) ne 'Math::BigFloat'; |
427
|
|
|
|
|
|
|
|
428
|
|
|
|
|
|
|
# taylor: x^2 x^4 x^6 |
429
|
|
|
|
|
|
|
# cosh = x + --- + --- + --- ... |
430
|
|
|
|
|
|
|
# 2! 4! 6! |
431
|
|
|
|
|
|
|
|
432
|
|
|
|
|
|
|
# difference for each term is thus x^2 and 1,2 |
433
|
|
|
|
|
|
|
|
434
|
0
|
|
|
|
|
0
|
my $cosh = Math::BigFloat->bone(); my $last = 0; |
|
0
|
|
|
|
|
0
|
|
435
|
0
|
|
|
|
|
0
|
my $x2 = $x*$x; |
436
|
0
|
|
|
|
|
0
|
my $over = $x2; my $below = Math::BigFloat->new(); my $factorial = Math::BigFloat->new(3); |
|
0
|
|
|
|
|
0
|
|
|
0
|
|
|
|
|
0
|
|
437
|
0
|
|
|
|
|
0
|
while ($cosh->bcmp($last)) # no $x-$last > $diff because bdiv() limit on accuracy |
438
|
|
|
|
|
|
|
{ |
439
|
0
|
|
|
|
|
0
|
$last = $cosh->copy(); |
440
|
0
|
|
|
|
|
0
|
$cosh += $over->copy()->bdiv($below,$d); |
441
|
0
|
|
|
|
|
0
|
$over *= $x2; # $x*$x |
442
|
0
|
|
|
|
|
0
|
$below *= $factorial; $factorial++; # n*(n+1) |
|
0
|
|
|
|
|
0
|
|
443
|
0
|
|
|
|
|
0
|
$below *= $factorial; $factorial++; |
|
0
|
|
|
|
|
0
|
|
444
|
|
|
|
|
|
|
} |
445
|
0
|
|
|
|
|
0
|
$cosh->bround($d-1); |
446
|
|
|
|
|
|
|
} |
447
|
|
|
|
|
|
|
|
448
|
|
|
|
|
|
|
sub arctan |
449
|
|
|
|
|
|
|
{ |
450
|
|
|
|
|
|
|
# calculate arcus tangens |
451
|
|
|
|
|
|
|
# first argument is x, so that result is arctan(x) |
452
|
|
|
|
|
|
|
# Second argument is accuracy (number of significant digits), it |
453
|
|
|
|
|
|
|
# stops when at least so much plus one digits are 'stable' and then |
454
|
|
|
|
|
|
|
# rounds it. Default is 42. |
455
|
8
|
50
|
|
8
|
1
|
6961
|
my $x = shift; $x = 0 if !defined $x; |
|
8
|
|
|
|
|
28
|
|
456
|
8
|
|
50
|
|
|
24
|
my $d = abs(shift || 42); $d = abs($d)+1; |
|
8
|
|
|
|
|
15
|
|
457
|
|
|
|
|
|
|
|
458
|
8
|
100
|
|
|
|
32
|
$x = Math::BigFloat->new($x) if ref($x) ne 'Math::BigFloat'; |
459
|
|
|
|
|
|
|
|
460
|
|
|
|
|
|
|
# taylor: x^3 x^5 x^7 x^9 |
461
|
|
|
|
|
|
|
# arctan = x - --- + --- - --- + --- ... |
462
|
|
|
|
|
|
|
# 3 5 7 9 |
463
|
|
|
|
|
|
|
|
464
|
|
|
|
|
|
|
# difference for each term is thus x^2 and 2: |
465
|
|
|
|
|
|
|
# 2 copy, 1 bmul, 1 badd, 1 bdiv |
466
|
|
|
|
|
|
|
|
467
|
8
|
|
|
|
|
699
|
my $arctan = $x->copy(); my $last = 0; |
|
8
|
|
|
|
|
228
|
|
468
|
8
|
|
|
|
|
25
|
my $x2 = $x*$x; |
469
|
8
|
|
|
|
|
2147
|
my $over = $x2*$x; my $below = Math::BigFloat->new(3); my $add = Math::BigFloat->new(2); |
|
8
|
|
|
|
|
2065
|
|
|
8
|
|
|
|
|
866
|
|
470
|
8
|
|
|
|
|
745
|
my $sign = 1; |
471
|
8
|
|
|
|
|
22
|
while ($arctan->bcmp($last)) # no $x-$last > $diff because bdiv() limit on A |
472
|
|
|
|
|
|
|
{ |
473
|
85
|
|
|
|
|
28169
|
$last = $arctan->copy(); |
474
|
85
|
100
|
|
|
|
2537
|
if ($sign == 0) |
475
|
|
|
|
|
|
|
{ |
476
|
42
|
|
|
|
|
81
|
$arctan += $over->copy()->bdiv($below,$d); |
477
|
|
|
|
|
|
|
} |
478
|
|
|
|
|
|
|
else |
479
|
|
|
|
|
|
|
{ |
480
|
43
|
|
|
|
|
89
|
$arctan -= $over->copy()->bdiv($below,$d); |
481
|
|
|
|
|
|
|
} |
482
|
85
|
|
|
|
|
123116
|
$sign = 1-$sign; # alternate |
483
|
85
|
|
|
|
|
214
|
$over *= $x2; # $x*$x |
484
|
85
|
|
|
|
|
20572
|
$below += $add; |
485
|
|
|
|
|
|
|
} |
486
|
8
|
|
|
|
|
2356
|
$arctan->bround($d-1); |
487
|
|
|
|
|
|
|
} |
488
|
|
|
|
|
|
|
|
489
|
|
|
|
|
|
|
sub arctanh |
490
|
|
|
|
|
|
|
{ |
491
|
|
|
|
|
|
|
# calculate arcus tangens hyperbolicus |
492
|
|
|
|
|
|
|
# first argument is x, so that result is arctanh(x) |
493
|
|
|
|
|
|
|
# Second argument is accuracy (number of significant digits), it |
494
|
|
|
|
|
|
|
# stops when at least so much plus one digits are 'stable' and then |
495
|
|
|
|
|
|
|
# rounds it. Default is 42. |
496
|
2
|
50
|
|
2
|
1
|
1889
|
my $x = shift; $x = 0 if !defined $x; |
|
2
|
|
|
|
|
7
|
|
497
|
2
|
|
50
|
|
|
9
|
my $d = abs(shift || 42); $d = abs($d)+1; |
|
2
|
|
|
|
|
4
|
|
498
|
|
|
|
|
|
|
|
499
|
2
|
50
|
|
|
|
10
|
$x = Math::BigFloat->new($x) if ref($x) ne 'Math::BigFloat'; |
500
|
|
|
|
|
|
|
|
501
|
|
|
|
|
|
|
# taylor: x^3 x^5 x^7 x^9 |
502
|
|
|
|
|
|
|
# arctanh = x + --- + --- + --- + --- + ... |
503
|
|
|
|
|
|
|
# 3 5 7 9 |
504
|
|
|
|
|
|
|
|
505
|
|
|
|
|
|
|
# difference for each term is thus x^2 and 2: |
506
|
|
|
|
|
|
|
# 2 copy, 1 bmul, 1 badd, 1 bdiv |
507
|
|
|
|
|
|
|
|
508
|
2
|
|
|
|
|
242
|
my $arctanh = $x->copy(); my $last = 0; |
|
2
|
|
|
|
|
55
|
|
509
|
2
|
|
|
|
|
7
|
my $x2 = $x*$x; |
510
|
2
|
|
|
|
|
281
|
my $over = $x2*$x; my $below = Math::BigFloat->new(3); my $add = Math::BigFloat->new(2); |
|
2
|
|
|
|
|
233
|
|
|
2
|
|
|
|
|
206
|
|
511
|
2
|
|
|
|
|
212
|
while ($arctanh->bcmp($last)) # no $x-$last > $diff because bdiv() limit on A |
512
|
|
|
|
|
|
|
{ |
513
|
0
|
|
|
|
|
0
|
$last = $arctanh->copy(); |
514
|
0
|
|
|
|
|
0
|
$arctanh += $over->copy()->bdiv($below,$d); |
515
|
0
|
|
|
|
|
0
|
$over *= $x2; # $x*$x |
516
|
0
|
|
|
|
|
0
|
$below += $add; |
517
|
|
|
|
|
|
|
} |
518
|
2
|
|
|
|
|
384
|
$arctanh->bround($d-1); |
519
|
|
|
|
|
|
|
} |
520
|
|
|
|
|
|
|
|
521
|
|
|
|
|
|
|
sub arcsin |
522
|
|
|
|
|
|
|
{ |
523
|
|
|
|
|
|
|
# calculate arcus sinus |
524
|
|
|
|
|
|
|
# first argument is x, so that result is arcsin(x) |
525
|
|
|
|
|
|
|
# Second argument is accuracy (number of significant digits), it |
526
|
|
|
|
|
|
|
# stops when at least so much plus one digits are 'stable' and then |
527
|
|
|
|
|
|
|
# rounds it. Default is 42. |
528
|
4
|
50
|
|
4
|
1
|
2973
|
my $x = shift; $x = 0 if !defined $x; |
|
4
|
|
|
|
|
23
|
|
529
|
4
|
|
100
|
|
|
17
|
my $d = abs(shift || 42); $d = abs($d)+1; |
|
4
|
|
|
|
|
7
|
|
530
|
|
|
|
|
|
|
|
531
|
4
|
50
|
|
|
|
19
|
$x = Math::BigFloat->new($x) if ref($x) ne 'Math::BigFloat'; |
532
|
|
|
|
|
|
|
|
533
|
|
|
|
|
|
|
# taylor: 1 * x^3 1 * 3 * x^5 1 * 3 * 5 * x^7 |
534
|
|
|
|
|
|
|
# arcsin = x + ------- + ----------- + --------------- + ... |
535
|
|
|
|
|
|
|
# 2 * 3 2 * 4 * 5 2 * 4 * 6 * 7 |
536
|
|
|
|
|
|
|
|
537
|
|
|
|
|
|
|
# difference for each term is thus x^2 and two muls (fac1, fac2): |
538
|
|
|
|
|
|
|
# 3 copy, 3 bmul, 1 bdiv, 3 badd |
539
|
|
|
|
|
|
|
|
540
|
4
|
|
|
|
|
559
|
my $arcsin = $x->copy(); my $last = 0; |
|
4
|
|
|
|
|
109
|
|
541
|
4
|
|
|
|
|
11
|
my $x2 = $x*$x; |
542
|
4
|
|
|
|
|
493
|
my $over = $x2*$x; my $below = Math::BigFloat->new(6); |
|
4
|
|
|
|
|
462
|
|
543
|
4
|
|
|
|
|
398
|
my $fac1 = Math::BigFloat->new(1); |
544
|
4
|
|
|
|
|
370
|
my $fac2 = Math::BigFloat->new(2); |
545
|
4
|
|
|
|
|
362
|
my $two = Math::BigFloat->new(2); |
546
|
4
|
|
|
|
|
365
|
while ($arcsin->bcmp($last)) # no $x-$last > $diff because bdiv() limit on A |
547
|
|
|
|
|
|
|
{ |
548
|
30
|
|
|
|
|
11444
|
$last = $arcsin->copy(); |
549
|
30
|
|
|
|
|
917
|
$arcsin += $over->copy()->bmul($fac1)->bdiv($below->copy->bmul($fac2),$d); |
550
|
30
|
|
|
|
|
51844
|
$over *= $x2; # $x*$x |
551
|
30
|
|
|
|
|
3548
|
$below += $one; |
552
|
30
|
|
|
|
|
9631
|
$fac1 += $two; |
553
|
30
|
|
|
|
|
4475
|
$fac2 += $two; |
554
|
|
|
|
|
|
|
} |
555
|
4
|
|
|
|
|
1006
|
$arcsin->bround($d-1); |
556
|
|
|
|
|
|
|
} |
557
|
|
|
|
|
|
|
|
558
|
|
|
|
|
|
|
sub arcsinh |
559
|
|
|
|
|
|
|
{ |
560
|
|
|
|
|
|
|
# calculate arcus sinus hyperbolicus |
561
|
|
|
|
|
|
|
# first argument is x, so that result is arcsinh(x) |
562
|
|
|
|
|
|
|
# Second argument is accuracy (number of significant digits), it |
563
|
|
|
|
|
|
|
# stops when at least so much plus one digits are 'stable' and then |
564
|
|
|
|
|
|
|
# rounds it. Default is 42. |
565
|
2
|
50
|
|
2
|
1
|
2136
|
my $x = shift; $x = 0 if !defined $x; |
|
2
|
|
|
|
|
8
|
|
566
|
2
|
|
50
|
|
|
9
|
my $d = abs(shift || 42); $d = abs($d)+1; |
|
2
|
|
|
|
|
3
|
|
567
|
|
|
|
|
|
|
|
568
|
2
|
50
|
|
|
|
12
|
$x = Math::BigFloat->new($x) if ref($x) ne 'Math::BigFloat'; |
569
|
|
|
|
|
|
|
|
570
|
|
|
|
|
|
|
# taylor: 1 * x^3 1 * 3 * x^5 1 * 3 * 5 * x^7 |
571
|
|
|
|
|
|
|
# arcsin = x - ------- + ----------- - --------------- + ... |
572
|
|
|
|
|
|
|
# 2 * 3 2 * 4 * 5 2 * 4 * 6 * 7 |
573
|
|
|
|
|
|
|
|
574
|
|
|
|
|
|
|
# difference for each term is thus x^2 and two muls (fac1, fac2): |
575
|
|
|
|
|
|
|
# 3 copy, 3 bmul, 1 bdiv, 3 badd |
576
|
|
|
|
|
|
|
|
577
|
2
|
|
|
|
|
260
|
my $arcsinh = $x->copy(); my $last = 0; |
|
2
|
|
|
|
|
56
|
|
578
|
2
|
|
|
|
|
8
|
my $x2 = $x*$x; my $sign = 0; |
|
2
|
|
|
|
|
240
|
|
579
|
2
|
|
|
|
|
5
|
my $over = $x2*$x; my $below = 6; |
|
2
|
|
|
|
|
221
|
|
580
|
2
|
|
|
|
|
7
|
my $fac1 = Math::BigInt->new(1); |
581
|
2
|
|
|
|
|
96
|
my $fac2 = Math::BigInt->new(2); |
582
|
2
|
|
|
|
|
73
|
while ($arcsinh ne $last) # no $x-$last > $diff because bdiv() limit on A |
583
|
|
|
|
|
|
|
{ |
584
|
0
|
|
|
|
|
0
|
$last = $arcsinh->copy(); |
585
|
0
|
0
|
|
|
|
0
|
if ($sign == 0) |
586
|
|
|
|
|
|
|
{ |
587
|
0
|
|
|
|
|
0
|
$arcsinh += $over->copy()->bmul( |
588
|
|
|
|
|
|
|
$fac1)->bdiv($below->copy->bmul($fac2),$d); |
589
|
|
|
|
|
|
|
} |
590
|
|
|
|
|
|
|
else |
591
|
|
|
|
|
|
|
{ |
592
|
0
|
|
|
|
|
0
|
$arcsinh -= $over->copy()->bmul( |
593
|
|
|
|
|
|
|
$fac1)->bdiv($below->copy->bmul($fac2),$d); |
594
|
|
|
|
|
|
|
} |
595
|
0
|
|
|
|
|
0
|
$over *= $x2; # $x*$x |
596
|
0
|
|
|
|
|
0
|
$below += $one; |
597
|
0
|
|
|
|
|
0
|
$fac1 += $two; |
598
|
0
|
|
|
|
|
0
|
$fac2 += $two; |
599
|
|
|
|
|
|
|
} |
600
|
2
|
|
|
|
|
64
|
$arcsinh->round($d-1); |
601
|
|
|
|
|
|
|
} |
602
|
|
|
|
|
|
|
|
603
|
|
|
|
|
|
|
sub log |
604
|
|
|
|
|
|
|
{ |
605
|
5
|
|
|
5
|
1
|
4237
|
my ($x,$base,$d) = @_; |
606
|
|
|
|
|
|
|
|
607
|
5
|
|
|
|
|
7
|
my $y; |
608
|
5
|
50
|
33
|
|
|
21
|
if (!ref($x) || !$x->isa('Math::BigFloat')) |
609
|
|
|
|
|
|
|
{ |
610
|
5
|
|
|
|
|
20
|
$y = Math::BigFloat->new($x); |
611
|
|
|
|
|
|
|
} |
612
|
|
|
|
|
|
|
else |
613
|
|
|
|
|
|
|
{ |
614
|
0
|
|
|
|
|
0
|
$y = $x->copy(); |
615
|
|
|
|
|
|
|
} |
616
|
5
|
|
|
|
|
979
|
$y->blog($base,$d); |
617
|
5
|
|
|
|
|
4863
|
$y; |
618
|
|
|
|
|
|
|
} |
619
|
|
|
|
|
|
|
|
620
|
|
|
|
|
|
|
sub pi |
621
|
|
|
|
|
|
|
{ |
622
|
|
|
|
|
|
|
# calculate PI (as suggested by Robert Creager) |
623
|
2
|
|
50
|
2
|
1
|
1870
|
my $digits = abs(shift || 1024); |
624
|
|
|
|
|
|
|
|
625
|
2
|
|
|
|
|
5
|
my $d = $digits+5; |
626
|
|
|
|
|
|
|
|
627
|
2
|
|
|
|
|
8
|
my $pi = $sixteen * arctan( scalar $fone->copy()->bdiv($five,$d), $d ) |
628
|
|
|
|
|
|
|
- $four * arctan( scalar $fone->copy()->bdiv($twothreenine,$d), $d); |
629
|
2
|
|
|
|
|
2679
|
$pi->bround($digits+1); # +1 for the "3." |
630
|
|
|
|
|
|
|
} |
631
|
|
|
|
|
|
|
|
632
|
|
|
|
|
|
|
1; |
633
|
|
|
|
|
|
|
|
634
|
|
|
|
|
|
|
__END__ |