1 package Math::BigFloat;
4 # Mike grinned. 'Two down, infinity to go' - Mike Nostrus in 'Before and After'
7 # The following hash values are internally used:
8 # _e : exponent (ref to $CALC object)
9 # _m : mantissa (ref to $CALC object)
11 # sign : +,-,+inf,-inf, or "NaN" if not a number
19 @ISA = qw(Exporter Math::BigInt);
22 # $_trap_inf/$_trap_nan are internal and should never be accessed from outside
23 use vars qw/$AUTOLOAD $accuracy $precision $div_scale $round_mode $rnd_mode
24 $upgrade $downgrade $_trap_nan $_trap_inf/;
25 my $class = "Math::BigFloat";
28 '<=>' => sub { my $rc = $_[2] ?
29 ref($_[0])->bcmp($_[1],$_[0]) :
30 ref($_[0])->bcmp($_[0],$_[1]);
31 $rc = 1 unless defined $rc;
34 # we need '>=' to get things like "1 >= NaN" right:
35 '>=' => sub { my $rc = $_[2] ?
36 ref($_[0])->bcmp($_[1],$_[0]) :
37 ref($_[0])->bcmp($_[0],$_[1]);
38 # if there was a NaN involved, return false
39 return '' unless defined $rc;
42 'int' => sub { $_[0]->as_number() }, # 'trunc' to bigint
45 ##############################################################################
46 # global constants, flags and assorted stuff
48 # the following are public, but their usage is not recommended. Use the
49 # accessor methods instead.
51 # class constants, use Class->constant_name() to access
52 $round_mode = 'even'; # one of 'even', 'odd', '+inf', '-inf', 'zero' or 'trunc'
59 # the package we are using for our private parts, defaults to:
60 # Math::BigInt->config()->{lib}
61 my $MBI = 'Math::BigInt::FastCalc';
63 # are NaNs ok? (otherwise it dies when encountering an NaN) set w/ config()
65 # the same for infinity
68 # constant for easier life
71 my $IMPORT = 0; # was import() called yet? used to make require work
73 # some digits of accuracy for blog(undef,10); which we use in blog() for speed
75 '2.3025850929940456840179914546843642076011014886287729760333279009675726097';
76 my $LOG_10_A = length($LOG_10)-1;
79 '0.6931471805599453094172321214581765680755001343602552541206800094933936220';
80 my $LOG_2_A = length($LOG_2)-1;
81 my $HALF = '0.5'; # made into an object if nec.
83 ##############################################################################
84 # the old code had $rnd_mode, so we need to support it, too
86 sub TIESCALAR { my ($class) = @_; bless \$round_mode, $class; }
87 sub FETCH { return $round_mode; }
88 sub STORE { $rnd_mode = $_[0]->round_mode($_[1]); }
92 # when someone set's $rnd_mode, we catch this and check the value to see
93 # whether it is valid or not.
94 $rnd_mode = 'even'; tie $rnd_mode, 'Math::BigFloat';
96 # we need both of them in this package:
97 *as_int = \&as_number;
100 ##############################################################################
103 # valid method aliases for AUTOLOAD
104 my %methods = map { $_ => 1 }
105 qw / fadd fsub fmul fdiv fround ffround fsqrt fmod fstr fsstr fpow fnorm
106 fint facmp fcmp fzero fnan finf finc fdec flog ffac fneg
107 fceil ffloor frsft flsft fone flog froot
109 # valid methods that can be handed up (for AUTOLOAD)
110 my %hand_ups = map { $_ => 1 }
111 qw / is_nan is_inf is_negative is_positive is_pos is_neg
112 accuracy precision div_scale round_mode fabs fnot
113 objectify upgrade downgrade
118 sub _method_alias { exists $methods{$_[0]||''}; }
119 sub _method_hand_up { exists $hand_ups{$_[0]||''}; }
122 ##############################################################################
127 # create a new BigFloat object from a string or another bigfloat object.
130 # sign => sign (+/-), or "NaN"
132 my ($class,$wanted,@r) = @_;
134 # avoid numify-calls by not using || on $wanted!
135 return $class->bzero() if !defined $wanted; # default to 0
136 return $wanted->copy() if UNIVERSAL::isa($wanted,'Math::BigFloat');
138 $class->import() if $IMPORT == 0; # make require work
140 my $self = {}; bless $self, $class;
141 # shortcut for bigints and its subclasses
142 if ((ref($wanted)) && UNIVERSAL::can( $wanted, "as_number"))
144 $self->{_m} = $wanted->as_number()->{value}; # get us a bigint copy
145 $self->{_e} = $MBI->_zero();
147 $self->{sign} = $wanted->sign();
148 return $self->bnorm();
150 # else: got a string or something maskerading as number (with overload)
152 # handle '+inf', '-inf' first
153 if ($wanted =~ /^[+-]?inf\z/)
155 return $downgrade->new($wanted) if $downgrade;
157 $self->{sign} = $wanted; # set a default sign for bstr()
158 return $self->binf($wanted);
161 # shortcut for simple forms like '12' that neither have trailing nor leading
163 if ($wanted =~ /^([+-]?)([1-9][0-9]*[1-9])$/)
165 $self->{_e} = $MBI->_zero();
167 $self->{sign} = $1 || '+';
168 $self->{_m} = $MBI->_new($2);
169 return $self->round(@r) if !$downgrade;
172 my ($mis,$miv,$mfv,$es,$ev) = Math::BigInt::_split($wanted);
178 Carp::croak ("$wanted is not a number initialized to $class");
181 return $downgrade->bnan() if $downgrade;
183 $self->{_e} = $MBI->_zero();
185 $self->{_m} = $MBI->_zero();
186 $self->{sign} = $nan;
190 # make integer from mantissa by adjusting exp, then convert to int
191 $self->{_e} = $MBI->_new($$ev); # exponent
192 $self->{_es} = $$es || '+';
193 my $mantissa = "$$miv$$mfv"; # create mant.
194 $mantissa =~ s/^0+(\d)/$1/; # strip leading zeros
195 $self->{_m} = $MBI->_new($mantissa); # create mant.
197 # 3.123E0 = 3123E-3, and 3.123E-2 => 3123E-5
198 if (CORE::length($$mfv) != 0)
200 my $len = $MBI->_new( CORE::length($$mfv));
201 ($self->{_e}, $self->{_es}) =
202 _e_sub ($self->{_e}, $len, $self->{_es}, '+');
204 # we can only have trailing zeros on the mantissa if $$mfv eq ''
207 # Use a regexp to count the trailing zeros in $$miv instead of _zeros()
208 # because that is faster, especially when _m is not stored in base 10.
209 my $zeros = 0; $zeros = CORE::length($1) if $$miv =~ /[1-9](0*)$/;
212 my $z = $MBI->_new($zeros);
213 # turn '120e2' into '12e3'
214 $MBI->_rsft ( $self->{_m}, $z, 10);
215 ($self->{_e}, $self->{_es}) =
216 _e_add ( $self->{_e}, $z, $self->{_es}, '+');
219 $self->{sign} = $$mis;
221 # for something like 0Ey, set y to 1, and -0 => +0
222 # Check $$miv for being '0' and $$mfv eq '', because otherwise _m could not
223 # have become 0. That's faster than to call $MBI->_is_zero().
224 $self->{sign} = '+', $self->{_e} = $MBI->_one()
225 if $$miv eq '0' and $$mfv eq '';
227 return $self->round(@r) if !$downgrade;
229 # if downgrade, inf, NaN or integers go down
231 if ($downgrade && $self->{_es} eq '+')
233 if ($MBI->_is_zero( $self->{_e} ))
235 return $downgrade->new($$mis . $MBI->_str( $self->{_m} ));
237 return $downgrade->new($self->bsstr());
239 $self->bnorm()->round(@r); # first normalize, then round
247 # if two arguments, the first one is the class to "swallow" subclasses
255 return unless ref($x); # only for objects
257 my $self = {}; bless $self,$c;
259 $self->{sign} = $x->{sign};
260 $self->{_es} = $x->{_es};
261 $self->{_m} = $MBI->_copy($x->{_m});
262 $self->{_e} = $MBI->_copy($x->{_e});
263 $self->{_a} = $x->{_a} if defined $x->{_a};
264 $self->{_p} = $x->{_p} if defined $x->{_p};
270 # used by parent class bone() to initialize number to NaN
276 my $class = ref($self);
277 Carp::croak ("Tried to set $self to NaN in $class\::_bnan()");
280 $IMPORT=1; # call our import only once
281 $self->{_m} = $MBI->_zero();
282 $self->{_e} = $MBI->_zero();
288 # used by parent class bone() to initialize number to +-inf
294 my $class = ref($self);
295 Carp::croak ("Tried to set $self to +-inf in $class\::_binf()");
298 $IMPORT=1; # call our import only once
299 $self->{_m} = $MBI->_zero();
300 $self->{_e} = $MBI->_zero();
306 # used by parent class bone() to initialize number to 1
308 $IMPORT=1; # call our import only once
309 $self->{_m} = $MBI->_one();
310 $self->{_e} = $MBI->_zero();
316 # used by parent class bone() to initialize number to 0
318 $IMPORT=1; # call our import only once
319 $self->{_m} = $MBI->_zero();
320 $self->{_e} = $MBI->_one();
326 my ($self,$class) = @_;
327 return if $class =~ /^Math::BigInt/; # we aren't one of these
328 UNIVERSAL::isa($self,$class);
333 # return (later set?) configuration data as hash ref
334 my $class = shift || 'Math::BigFloat';
336 my $cfg = $class->SUPER::config(@_);
338 # now we need only to override the ones that are different from our parent
339 $cfg->{class} = $class;
344 ##############################################################################
345 # string conversation
349 # (ref to BFLOAT or num_str ) return num_str
350 # Convert number from internal format to (non-scientific) string format.
351 # internal format is always normalized (no leading zeros, "-0" => "+0")
352 my ($self,$x) = ref($_[0]) ? (undef,$_[0]) : objectify(1,@_);
354 if ($x->{sign} !~ /^[+-]$/)
356 return $x->{sign} unless $x->{sign} eq '+inf'; # -inf, NaN
360 my $es = '0'; my $len = 1; my $cad = 0; my $dot = '.';
363 my $not_zero = !($x->{sign} eq '+' && $MBI->_is_zero($x->{_m}));
366 $es = $MBI->_str($x->{_m});
367 $len = CORE::length($es);
368 my $e = $MBI->_num($x->{_e});
369 $e = -$e if $x->{_es} eq '-';
373 # if _e is bigger than a scalar, the following will blow your memory
376 my $r = abs($e) - $len;
377 $es = '0.'. ('0' x $r) . $es; $cad = -($len+$r);
381 substr($es,$e,0) = '.'; $cad = $MBI->_num($x->{_e});
382 $cad = -$cad if $x->{_es} eq '-';
388 $es .= '0' x $e; $len += $e; $cad = 0;
392 $es = '-'.$es if $x->{sign} eq '-';
393 # if set accuracy or precision, pad with zeros on the right side
394 if ((defined $x->{_a}) && ($not_zero))
396 # 123400 => 6, 0.1234 => 4, 0.001234 => 4
397 my $zeros = $x->{_a} - $cad; # cad == 0 => 12340
398 $zeros = $x->{_a} - $len if $cad != $len;
399 $es .= $dot.'0' x $zeros if $zeros > 0;
401 elsif ((($x->{_p} || 0) < 0))
403 # 123400 => 6, 0.1234 => 4, 0.001234 => 6
404 my $zeros = -$x->{_p} + $cad;
405 $es .= $dot.'0' x $zeros if $zeros > 0;
412 # (ref to BFLOAT or num_str ) return num_str
413 # Convert number from internal format to scientific string format.
414 # internal format is always normalized (no leading zeros, "-0E0" => "+0E0")
415 my ($self,$x) = ref($_[0]) ? (undef,$_[0]) : objectify(1,@_);
417 if ($x->{sign} !~ /^[+-]$/)
419 return $x->{sign} unless $x->{sign} eq '+inf'; # -inf, NaN
422 my $sep = 'e'.$x->{_es};
423 my $sign = $x->{sign}; $sign = '' if $sign eq '+';
424 $sign . $MBI->_str($x->{_m}) . $sep . $MBI->_str($x->{_e});
429 # Make a number from a BigFloat object
430 # simple return a string and let Perl's atoi()/atof() handle the rest
431 my ($self,$x) = ref($_[0]) ? (undef,$_[0]) : objectify(1,@_);
435 ##############################################################################
436 # public stuff (usually prefixed with "b")
440 # (BINT or num_str) return BINT
441 # negate number or make a negated number from string
442 my ($self,$x) = ref($_[0]) ? (undef,$_[0]) : objectify(1,@_);
444 return $x if $x->modify('bneg');
446 # for +0 dont negate (to have always normalized +0). Does nothing for 'NaN'
447 $x->{sign} =~ tr/+-/-+/ unless ($x->{sign} eq '+' && $MBI->_is_zero($x->{_m}));
452 # XXX TODO this must be overwritten and return NaN for non-integer values
453 # band(), bior(), bxor(), too
456 # $class->SUPER::bnot($class,@_);
461 # Compares 2 values. Returns one of undef, <0, =0, >0. (suitable for sort)
464 my ($self,$x,$y) = (ref($_[0]),@_);
465 # objectify is costly, so avoid it
466 if ((!ref($_[0])) || (ref($_[0]) ne ref($_[1])))
468 ($self,$x,$y) = objectify(2,@_);
471 return $upgrade->bcmp($x,$y) if defined $upgrade &&
472 ((!$x->isa($self)) || (!$y->isa($self)));
474 if (($x->{sign} !~ /^[+-]$/) || ($y->{sign} !~ /^[+-]$/))
476 # handle +-inf and NaN
477 return undef if (($x->{sign} eq $nan) || ($y->{sign} eq $nan));
478 return 0 if ($x->{sign} eq $y->{sign}) && ($x->{sign} =~ /^[+-]inf$/);
479 return +1 if $x->{sign} eq '+inf';
480 return -1 if $x->{sign} eq '-inf';
481 return -1 if $y->{sign} eq '+inf';
485 # check sign for speed first
486 return 1 if $x->{sign} eq '+' && $y->{sign} eq '-'; # does also 0 <=> -y
487 return -1 if $x->{sign} eq '-' && $y->{sign} eq '+'; # does also -x <=> 0
490 my $xz = $x->is_zero();
491 my $yz = $y->is_zero();
492 return 0 if $xz && $yz; # 0 <=> 0
493 return -1 if $xz && $y->{sign} eq '+'; # 0 <=> +y
494 return 1 if $yz && $x->{sign} eq '+'; # +x <=> 0
496 # adjust so that exponents are equal
497 my $lxm = $MBI->_len($x->{_m});
498 my $lym = $MBI->_len($y->{_m});
499 # the numify somewhat limits our length, but makes it much faster
500 my ($xes,$yes) = (1,1);
501 $xes = -1 if $x->{_es} ne '+';
502 $yes = -1 if $y->{_es} ne '+';
503 my $lx = $lxm + $xes * $MBI->_num($x->{_e});
504 my $ly = $lym + $yes * $MBI->_num($y->{_e});
505 my $l = $lx - $ly; $l = -$l if $x->{sign} eq '-';
506 return $l <=> 0 if $l != 0;
508 # lengths (corrected by exponent) are equal
509 # so make mantissa equal length by padding with zero (shift left)
510 my $diff = $lxm - $lym;
511 my $xm = $x->{_m}; # not yet copy it
515 $ym = $MBI->_copy($y->{_m});
516 $ym = $MBI->_lsft($ym, $MBI->_new($diff), 10);
520 $xm = $MBI->_copy($x->{_m});
521 $xm = $MBI->_lsft($xm, $MBI->_new(-$diff), 10);
523 my $rc = $MBI->_acmp($xm,$ym);
524 $rc = -$rc if $x->{sign} eq '-'; # -124 < -123
530 # Compares 2 values, ignoring their signs.
531 # Returns one of undef, <0, =0, >0. (suitable for sort)
534 my ($self,$x,$y) = (ref($_[0]),@_);
535 # objectify is costly, so avoid it
536 if ((!ref($_[0])) || (ref($_[0]) ne ref($_[1])))
538 ($self,$x,$y) = objectify(2,@_);
541 return $upgrade->bacmp($x,$y) if defined $upgrade &&
542 ((!$x->isa($self)) || (!$y->isa($self)));
544 # handle +-inf and NaN's
545 if ($x->{sign} !~ /^[+-]$/ || $y->{sign} !~ /^[+-]$/)
547 return undef if (($x->{sign} eq $nan) || ($y->{sign} eq $nan));
548 return 0 if ($x->is_inf() && $y->is_inf());
549 return 1 if ($x->is_inf() && !$y->is_inf());
554 my $xz = $x->is_zero();
555 my $yz = $y->is_zero();
556 return 0 if $xz && $yz; # 0 <=> 0
557 return -1 if $xz && !$yz; # 0 <=> +y
558 return 1 if $yz && !$xz; # +x <=> 0
560 # adjust so that exponents are equal
561 my $lxm = $MBI->_len($x->{_m});
562 my $lym = $MBI->_len($y->{_m});
563 my ($xes,$yes) = (1,1);
564 $xes = -1 if $x->{_es} ne '+';
565 $yes = -1 if $y->{_es} ne '+';
566 # the numify somewhat limits our length, but makes it much faster
567 my $lx = $lxm + $xes * $MBI->_num($x->{_e});
568 my $ly = $lym + $yes * $MBI->_num($y->{_e});
570 return $l <=> 0 if $l != 0;
572 # lengths (corrected by exponent) are equal
573 # so make mantissa equal-length by padding with zero (shift left)
574 my $diff = $lxm - $lym;
575 my $xm = $x->{_m}; # not yet copy it
579 $ym = $MBI->_copy($y->{_m});
580 $ym = $MBI->_lsft($ym, $MBI->_new($diff), 10);
584 $xm = $MBI->_copy($x->{_m});
585 $xm = $MBI->_lsft($xm, $MBI->_new(-$diff), 10);
587 $MBI->_acmp($xm,$ym);
592 # add second arg (BFLOAT or string) to first (BFLOAT) (modifies first)
593 # return result as BFLOAT
596 my ($self,$x,$y,$a,$p,$r) = (ref($_[0]),@_);
597 # objectify is costly, so avoid it
598 if ((!ref($_[0])) || (ref($_[0]) ne ref($_[1])))
600 ($self,$x,$y,$a,$p,$r) = objectify(2,@_);
603 # inf and NaN handling
604 if (($x->{sign} !~ /^[+-]$/) || ($y->{sign} !~ /^[+-]$/))
607 return $x->bnan() if (($x->{sign} eq $nan) || ($y->{sign} eq $nan));
609 if (($x->{sign} =~ /^[+-]inf$/) && ($y->{sign} =~ /^[+-]inf$/))
611 # +inf++inf or -inf+-inf => same, rest is NaN
612 return $x if $x->{sign} eq $y->{sign};
615 # +-inf + something => +inf; something +-inf => +-inf
616 $x->{sign} = $y->{sign}, return $x if $y->{sign} =~ /^[+-]inf$/;
620 return $upgrade->badd($x,$y,$a,$p,$r) if defined $upgrade &&
621 ((!$x->isa($self)) || (!$y->isa($self)));
623 # speed: no add for 0+y or x+0
624 return $x->bround($a,$p,$r) if $y->is_zero(); # x+0
625 if ($x->is_zero()) # 0+y
627 # make copy, clobbering up x (modify in place!)
628 $x->{_e} = $MBI->_copy($y->{_e});
629 $x->{_es} = $y->{_es};
630 $x->{_m} = $MBI->_copy($y->{_m});
631 $x->{sign} = $y->{sign} || $nan;
632 return $x->round($a,$p,$r,$y);
635 # take lower of the two e's and adapt m1 to it to match m2
637 $e = $MBI->_zero() if !defined $e; # if no BFLOAT?
638 $e = $MBI->_copy($e); # make copy (didn't do it yet)
642 ($e,$es) = _e_sub($e, $x->{_e}, $y->{_es} || '+', $x->{_es});
644 my $add = $MBI->_copy($y->{_m});
646 if ($es eq '-') # < 0
648 $MBI->_lsft( $x->{_m}, $e, 10);
649 ($x->{_e},$x->{_es}) = _e_add($x->{_e}, $e, $x->{_es}, $es);
651 elsif (!$MBI->_is_zero($e)) # > 0
653 $MBI->_lsft($add, $e, 10);
655 # else: both e are the same, so just leave them
657 if ($x->{sign} eq $y->{sign})
660 $x->{_m} = $MBI->_add($x->{_m}, $add);
664 ($x->{_m}, $x->{sign}) =
665 _e_add($x->{_m}, $add, $x->{sign}, $y->{sign});
668 # delete trailing zeros, then round
669 $x->bnorm()->round($a,$p,$r,$y);
672 # sub bsub is inherited from Math::BigInt!
676 # increment arg by one
677 my ($self,$x,@r) = ref($_[0]) ? (ref($_[0]),@_) : objectify(1,@_);
679 if ($x->{_es} eq '-')
681 return $x->badd($self->bone(),@r); # digits after dot
684 if (!$MBI->_is_zero($x->{_e})) # _e == 0 for NaN, inf, -inf
686 # 1e2 => 100, so after the shift below _m has a '0' as last digit
687 $x->{_m} = $MBI->_lsft($x->{_m}, $x->{_e},10); # 1e2 => 100
688 $x->{_e} = $MBI->_zero(); # normalize
690 # we know that the last digit of $x will be '1' or '9', depending on the
694 if ($x->{sign} eq '+')
696 $MBI->_inc($x->{_m});
697 return $x->bnorm()->bround(@r);
699 elsif ($x->{sign} eq '-')
701 $MBI->_dec($x->{_m});
702 $x->{sign} = '+' if $MBI->_is_zero($x->{_m}); # -1 +1 => -0 => +0
703 return $x->bnorm()->bround(@r);
705 # inf, nan handling etc
706 $x->badd($self->bone(),@r); # badd() does round
711 # decrement arg by one
712 my ($self,$x,@r) = ref($_[0]) ? (ref($_[0]),@_) : objectify(1,@_);
714 if ($x->{_es} eq '-')
716 return $x->badd($self->bone('-'),@r); # digits after dot
719 if (!$MBI->_is_zero($x->{_e}))
721 $x->{_m} = $MBI->_lsft($x->{_m}, $x->{_e},10); # 1e2 => 100
722 $x->{_e} = $MBI->_zero(); # normalize
726 my $zero = $x->is_zero();
728 if (($x->{sign} eq '-') || $zero)
730 $MBI->_inc($x->{_m});
731 $x->{sign} = '-' if $zero; # 0 => 1 => -1
732 $x->{sign} = '+' if $MBI->_is_zero($x->{_m}); # -1 +1 => -0 => +0
733 return $x->bnorm()->round(@r);
736 elsif ($x->{sign} eq '+')
738 $MBI->_dec($x->{_m});
739 return $x->bnorm()->round(@r);
741 # inf, nan handling etc
742 $x->badd($self->bone('-'),@r); # does round
749 my ($self,$x,$base,$a,$p,$r) = ref($_[0]) ? (ref($_[0]),@_) : objectify(1,@_);
751 # $base > 0, $base != 1; if $base == undef default to $base == e
754 # we need to limit the accuracy to protect against overflow
757 ($x,@params) = $x->_find_round_parameters($a,$p,$r);
759 # also takes care of the "error in _find_round_parameters?" case
760 return $x->bnan() if $x->{sign} ne '+' || $x->is_zero();
762 # no rounding at all, so must use fallback
763 if (scalar @params == 0)
765 # simulate old behaviour
766 $params[0] = $self->div_scale(); # and round to it as accuracy
767 $params[1] = undef; # P = undef
768 $scale = $params[0]+4; # at least four more for proper round
769 $params[2] = $r; # round mode by caller or undef
770 $fallback = 1; # to clear a/p afterwards
774 # the 4 below is empirical, and there might be cases where it is not
776 $scale = abs($params[0] || $params[1]) + 4; # take whatever is defined
779 return $x->bzero(@params) if $x->is_one();
780 # base not defined => base == Euler's constant e
783 # make object, since we don't feed it through objectify() to still get the
784 # case of $base == undef
785 $base = $self->new($base) unless ref($base);
786 # $base > 0; $base != 1
787 return $x->bnan() if $base->is_zero() || $base->is_one() ||
788 $base->{sign} ne '+';
789 # if $x == $base, we know the result must be 1.0
790 if ($x->bcmp($base) == 0)
792 $x->bone('+',@params);
795 # clear a/p after round, since user did not request it
796 delete $x->{_a}; delete $x->{_p};
802 # when user set globals, they would interfere with our calculation, so
803 # disable them and later re-enable them
805 my $abr = "$self\::accuracy"; my $ab = $$abr; $$abr = undef;
806 my $pbr = "$self\::precision"; my $pb = $$pbr; $$pbr = undef;
807 # we also need to disable any set A or P on $x (_find_round_parameters took
808 # them already into account), since these would interfere, too
809 delete $x->{_a}; delete $x->{_p};
810 # need to disable $upgrade in BigInt, to avoid deep recursion
811 local $Math::BigInt::upgrade = undef;
812 local $Math::BigFloat::downgrade = undef;
814 # upgrade $x if $x is not a BigFloat (handle BigInt input)
815 if (!$x->isa('Math::BigFloat'))
817 $x = Math::BigFloat->new($x);
823 # If the base is defined and an integer, try to calculate integer result
824 # first. This is very fast, and in case the real result was found, we can
826 if (defined $base && $base->is_int() && $x->is_int())
828 my $i = $MBI->_copy( $x->{_m} );
829 $MBI->_lsft( $i, $x->{_e}, 10 ) unless $MBI->_is_zero($x->{_e});
830 my $int = Math::BigInt->bzero();
832 $int->blog($base->as_number());
834 if ($base->as_number()->bpow($int) == $x)
836 # found result, return it
837 $x->{_m} = $int->{value};
838 $x->{_e} = $MBI->_zero();
847 # first calculate the log to base e (using reduction by 10 (and probably 2))
848 $self->_log_10($x,$scale);
850 # and if a different base was requested, convert it
853 $base = Math::BigFloat->new($base) unless $base->isa('Math::BigFloat');
854 # not ln, but some other base (don't modify $base)
855 $x->bdiv( $base->copy()->blog(undef,$scale), $scale );
859 # shortcut to not run through _find_round_parameters again
860 if (defined $params[0])
862 $x->bround($params[0],$params[2]); # then round accordingly
866 $x->bfround($params[1],$params[2]); # then round accordingly
870 # clear a/p after round, since user did not request it
871 delete $x->{_a}; delete $x->{_p};
874 $$abr = $ab; $$pbr = $pb;
881 # internal log function to calculate ln() based on Taylor series.
882 # Modifies $x in place.
883 my ($self,$x,$scale) = @_;
885 # in case of $x == 1, result is 0
886 return $x->bzero() if $x->is_one();
888 # http://www.efunda.com/math/taylor_series/logarithmic.cfm?search_string=log
892 # Taylor: | u 1 u^3 1 u^5 |
893 # ln (x) = 2 | --- + - * --- + - * --- + ... | x > 0
894 # |_ v 3 v^3 5 v^5 _|
896 # This takes much more steps to calculate the result and is thus not used
899 # Taylor: | u 1 u^2 1 u^3 |
900 # ln (x) = 2 | --- + - * --- + - * --- + ... | x > 1/2
901 # |_ x 2 x^2 3 x^3 _|
903 my ($limit,$v,$u,$below,$factor,$two,$next,$over,$f);
905 $v = $x->copy(); $v->binc(); # v = x+1
906 $x->bdec(); $u = $x->copy(); # u = x-1; x = x-1
907 $x->bdiv($v,$scale); # first term: u/v
910 $u *= $u; $v *= $v; # u^2, v^2
911 $below->bmul($v); # u^3, v^3
913 $factor = $self->new(3); $f = $self->new(2);
915 my $steps = 0 if DEBUG;
916 $limit = $self->new("1E-". ($scale-1));
919 # we calculate the next term, and add it to the last
920 # when the next term is below our limit, it won't affect the outcome
921 # anymore, so we stop
923 # calculating the next term simple from over/below will result in quite
924 # a time hog if the input has many digits, since over and below will
925 # accumulate more and more digits, and the result will also have many
926 # digits, but in the end it is rounded to $scale digits anyway. So if we
927 # round $over and $below first, we save a lot of time for the division
928 # (not with log(1.2345), but try log (123**123) to see what I mean. This
929 # can introduce a rounding error if the division result would be f.i.
930 # 0.1234500000001 and we round it to 5 digits it would become 0.12346, but
931 # if we truncated $over and $below we might get 0.12345. Does this matter
932 # for the end result? So we give $over and $below 4 more digits to be
933 # on the safe side (unscientific error handling as usual... :+D
935 $next = $over->copy->bround($scale+4)->bdiv(
936 $below->copy->bmul($factor)->bround($scale+4),
940 ## $next = $over->copy()->bdiv($below->copy()->bmul($factor),$scale);
942 last if $next->bacmp($limit) <= 0;
944 delete $next->{_a}; delete $next->{_p};
946 # calculate things for the next term
947 $over *= $u; $below *= $v; $factor->badd($f);
950 $steps++; print "step $steps = $x\n" if $steps % 10 == 0;
953 $x->bmul($f); # $x *= 2
954 print "took $steps steps\n" if DEBUG;
959 # Internal log function based on reducing input to the range of 0.1 .. 9.99
960 # and then "correcting" the result to the proper one. Modifies $x in place.
961 my ($self,$x,$scale) = @_;
963 # taking blog() from numbers greater than 10 takes a *very long* time, so we
964 # break the computation down into parts based on the observation that:
965 # blog(x*y) = blog(x) + blog(y)
966 # We set $y here to multiples of 10 so that $x is below 1 (the smaller $x is
967 # the faster it get's, especially because 2*$x takes about 10 times as long,
968 # so by dividing $x by 10 we make it at least factor 100 faster...)
970 # The same observation is valid for numbers smaller than 0.1 (e.g. computing
971 # log(1) is fastest, and the farther away we get from 1, the longer it takes)
972 # so we also 'break' this down by multiplying $x with 10 and subtract the
973 # log(10) afterwards to get the correct result.
975 # calculate nr of digits before dot
976 my $dbd = $MBI->_num($x->{_e});
977 $dbd = -$dbd if $x->{_es} eq '-';
978 $dbd += $MBI->_len($x->{_m});
980 # more than one digit (e.g. at least 10), but *not* exactly 10 to avoid
983 my $calc = 1; # do some calculation?
985 # disable the shortcut for 10, since we need log(10) and this would recurse
987 if ($x->{_es} eq '+' && $MBI->_is_one($x->{_e}) && $MBI->_is_one($x->{_m}))
989 $dbd = 0; # disable shortcut
990 # we can use the cached value in these cases
991 if ($scale <= $LOG_10_A)
993 $x->bzero(); $x->badd($LOG_10);
994 $calc = 0; # no need to calc, but round
999 # disable the shortcut for 2, since we maybe have it cached
1000 if (($MBI->_is_zero($x->{_e}) && $MBI->_is_two($x->{_m})))
1002 $dbd = 0; # disable shortcut
1003 # we can use the cached value in these cases
1004 if ($scale <= $LOG_2_A)
1006 $x->bzero(); $x->badd($LOG_2);
1007 $calc = 0; # no need to calc, but round
1012 # if $x = 0.1, we know the result must be 0-log(10)
1013 if ($calc != 0 && $x->{_es} eq '-' && $MBI->_is_one($x->{_e}) &&
1014 $MBI->_is_one($x->{_m}))
1016 $dbd = 0; # disable shortcut
1017 # we can use the cached value in these cases
1018 if ($scale <= $LOG_10_A)
1020 $x->bzero(); $x->bsub($LOG_10);
1021 $calc = 0; # no need to calc, but round
1025 return if $calc == 0; # already have the result
1027 # default: these correction factors are undef and thus not used
1028 my $l_10; # value of ln(10) to A of $scale
1029 my $l_2; # value of ln(2) to A of $scale
1031 # $x == 2 => 1, $x == 13 => 2, $x == 0.1 => 0, $x == 0.01 => -1
1032 # so don't do this shortcut for 1 or 0
1033 if (($dbd > 1) || ($dbd < 0))
1035 # convert our cached value to an object if not already (avoid doing this
1036 # at import() time, since not everybody needs this)
1037 $LOG_10 = $self->new($LOG_10,undef,undef) unless ref $LOG_10;
1039 #print "x = $x, dbd = $dbd, calc = $calc\n";
1040 # got more than one digit before the dot, or more than one zero after the
1042 # log(123) == log(1.23) + log(10) * 2
1043 # log(0.0123) == log(1.23) - log(10) * 2
1045 if ($scale <= $LOG_10_A)
1048 $l_10 = $LOG_10->copy(); # copy for mul
1052 # else: slower, compute it (but don't cache it, because it could be big)
1053 # also disable downgrade for this code path
1054 local $Math::BigFloat::downgrade = undef;
1055 $l_10 = $self->new(10)->blog(undef,$scale); # scale+4, actually
1057 $dbd-- if ($dbd > 1); # 20 => dbd=2, so make it dbd=1
1058 $l_10->bmul( $self->new($dbd)); # log(10) * (digits_before_dot-1)
1065 ($x->{_e}, $x->{_es}) =
1066 _e_sub( $x->{_e}, $MBI->_new($dbd), $x->{_es}, $dbd_sign); # 123 => 1.23
1070 # Now: 0.1 <= $x < 10 (and possible correction in l_10)
1072 ### Since $x in the range 0.5 .. 1.5 is MUCH faster, we do a repeated div
1073 ### or mul by 2 (maximum times 3, since x < 10 and x > 0.1)
1075 $HALF = $self->new($HALF) unless ref($HALF);
1077 my $twos = 0; # default: none (0 times)
1078 my $two = $self->new(2);
1079 while ($x->bacmp($HALF) <= 0)
1081 $twos--; $x->bmul($two);
1083 while ($x->bacmp($two) >= 0)
1085 $twos++; $x->bdiv($two,$scale+4); # keep all digits
1087 # $twos > 0 => did mul 2, < 0 => did div 2 (never both)
1088 # calculate correction factor based on ln(2)
1091 $LOG_2 = $self->new($LOG_2,undef,undef) unless ref $LOG_2;
1092 if ($scale <= $LOG_2_A)
1095 $l_2 = $LOG_2->copy(); # copy for mul
1099 # else: slower, compute it (but don't cache it, because it could be big)
1100 # also disable downgrade for this code path
1101 local $Math::BigFloat::downgrade = undef;
1102 $l_2 = $two->blog(undef,$scale); # scale+4, actually
1104 $l_2->bmul($twos); # * -2 => subtract, * 2 => add
1107 $self->_log($x,$scale); # need to do the "normal" way
1108 $x->badd($l_10) if defined $l_10; # correct it by ln(10)
1109 $x->badd($l_2) if defined $l_2; # and maybe by ln(2)
1110 # all done, $x contains now the result
1115 # (BFLOAT or num_str, BFLOAT or num_str) return BFLOAT
1116 # does not modify arguments, but returns new object
1117 # Lowest Common Multiplicator
1119 my ($self,@arg) = objectify(0,@_);
1120 my $x = $self->new(shift @arg);
1121 while (@arg) { $x = Math::BigInt::__lcm($x,shift @arg); }
1127 # (BINT or num_str, BINT or num_str) return BINT
1128 # does not modify arguments, but returns new object
1131 $y = __PACKAGE__->new($y) if !ref($y);
1133 my $x = $y->copy()->babs(); # keep arguments
1135 return $x->bnan() if $x->{sign} !~ /^[+-]$/ # x NaN?
1136 || !$x->is_int(); # only for integers now
1140 my $t = shift; $t = $self->new($t) if !ref($t);
1141 $y = $t->copy()->babs();
1143 return $x->bnan() if $y->{sign} !~ /^[+-]$/ # y NaN?
1144 || !$y->is_int(); # only for integers now
1146 # greatest common divisor
1147 while (! $y->is_zero())
1149 ($x,$y) = ($y->copy(), $x->copy()->bmod($y));
1152 last if $x->is_one();
1157 ##############################################################################
1161 # Internal helper sub to take two positive integers and their signs and
1162 # then add them. Input ($CALC,$CALC,('+'|'-'),('+'|'-')),
1163 # output ($CALC,('+'|'-'))
1164 my ($x,$y,$xs,$ys) = @_;
1166 # if the signs are equal we can add them (-5 + -3 => -(5 + 3) => -8)
1169 $x = $MBI->_add ($x, $y ); # a+b
1170 # the sign follows $xs
1174 my $a = $MBI->_acmp($x,$y);
1177 $x = $MBI->_sub ($x , $y); # abs sub
1181 $x = $MBI->_zero(); # result is 0
1186 $x = $MBI->_sub ( $y, $x, 1 ); # abs sub
1194 # Internal helper sub to take two positive integers and their signs and
1195 # then subtract them. Input ($CALC,$CALC,('+'|'-'),('+'|'-')),
1196 # output ($CALC,('+'|'-'))
1197 my ($x,$y,$xs,$ys) = @_;
1201 _e_add($x,$y,$xs,$ys); # call add (does subtract now)
1204 ###############################################################################
1205 # is_foo methods (is_negative, is_positive are inherited from BigInt)
1209 # return true if arg (BFLOAT or num_str) is an integer
1210 my ($self,$x) = ref($_[0]) ? (undef,$_[0]) : objectify(1,@_);
1212 return 1 if ($x->{sign} =~ /^[+-]$/) && # NaN and +-inf aren't
1213 $x->{_es} eq '+'; # 1e-1 => no integer
1219 # return true if arg (BFLOAT or num_str) is zero
1220 my ($self,$x) = ref($_[0]) ? (undef,$_[0]) : objectify(1,@_);
1222 return 1 if $x->{sign} eq '+' && $MBI->_is_zero($x->{_m});
1228 # return true if arg (BFLOAT or num_str) is +1 or -1 if signis given
1229 my ($self,$x,$sign) = ref($_[0]) ? (undef,@_) : objectify(1,@_);
1231 $sign = '+' if !defined $sign || $sign ne '-';
1233 if ($x->{sign} eq $sign &&
1234 $MBI->_is_zero($x->{_e}) && $MBI->_is_one($x->{_m}));
1240 # return true if arg (BFLOAT or num_str) is odd or false if even
1241 my ($self,$x) = ref($_[0]) ? (undef,$_[0]) : objectify(1,@_);
1243 return 1 if ($x->{sign} =~ /^[+-]$/) && # NaN & +-inf aren't
1244 ($MBI->_is_zero($x->{_e}) && $MBI->_is_odd($x->{_m}));
1250 # return true if arg (BINT or num_str) is even or false if odd
1251 my ($self,$x) = ref($_[0]) ? (undef,$_[0]) : objectify(1,@_);
1253 return 0 if $x->{sign} !~ /^[+-]$/; # NaN & +-inf aren't
1254 return 1 if ($x->{_es} eq '+' # 123.45 is never
1255 && $MBI->_is_even($x->{_m})); # but 1200 is
1261 # multiply two numbers -- stolen from Knuth Vol 2 pg 233
1262 # (BINT or num_str, BINT or num_str) return BINT
1265 my ($self,$x,$y,$a,$p,$r) = (ref($_[0]),@_);
1266 # objectify is costly, so avoid it
1267 if ((!ref($_[0])) || (ref($_[0]) ne ref($_[1])))
1269 ($self,$x,$y,$a,$p,$r) = objectify(2,@_);
1272 return $x->bnan() if (($x->{sign} eq $nan) || ($y->{sign} eq $nan));
1275 if (($x->{sign} =~ /^[+-]inf$/) || ($y->{sign} =~ /^[+-]inf$/))
1277 return $x->bnan() if $x->is_zero() || $y->is_zero();
1278 # result will always be +-inf:
1279 # +inf * +/+inf => +inf, -inf * -/-inf => +inf
1280 # +inf * -/-inf => -inf, -inf * +/+inf => -inf
1281 return $x->binf() if ($x->{sign} =~ /^\+/ && $y->{sign} =~ /^\+/);
1282 return $x->binf() if ($x->{sign} =~ /^-/ && $y->{sign} =~ /^-/);
1283 return $x->binf('-');
1286 return $x->bzero() if $x->is_zero() || $y->is_zero();
1288 return $upgrade->bmul($x,$y,$a,$p,$r) if defined $upgrade &&
1289 ((!$x->isa($self)) || (!$y->isa($self)));
1291 # aEb * cEd = (a*c)E(b+d)
1292 $MBI->_mul($x->{_m},$y->{_m});
1293 ($x->{_e}, $x->{_es}) = _e_add($x->{_e}, $y->{_e}, $x->{_es}, $y->{_es});
1296 $x->{sign} = $x->{sign} ne $y->{sign} ? '-' : '+';
1297 return $x->bnorm()->round($a,$p,$r,$y);
1302 # (dividend: BFLOAT or num_str, divisor: BFLOAT or num_str) return
1303 # (BFLOAT,BFLOAT) (quo,rem) or BFLOAT (only rem)
1306 my ($self,$x,$y,$a,$p,$r) = (ref($_[0]),@_);
1307 # objectify is costly, so avoid it
1308 if ((!ref($_[0])) || (ref($_[0]) ne ref($_[1])))
1310 ($self,$x,$y,$a,$p,$r) = objectify(2,@_);
1313 return $self->_div_inf($x,$y)
1314 if (($x->{sign} !~ /^[+-]$/) || ($y->{sign} !~ /^[+-]$/) || $y->is_zero());
1316 # x== 0 # also: or y == 1 or y == -1
1317 return wantarray ? ($x,$self->bzero()) : $x if $x->is_zero();
1320 return $upgrade->bdiv($upgrade->new($x),$y,$a,$p,$r) if defined $upgrade;
1322 # we need to limit the accuracy to protect against overflow
1324 my (@params,$scale);
1325 ($x,@params) = $x->_find_round_parameters($a,$p,$r,$y);
1327 return $x if $x->is_nan(); # error in _find_round_parameters?
1329 # no rounding at all, so must use fallback
1330 if (scalar @params == 0)
1332 # simulate old behaviour
1333 $params[0] = $self->div_scale(); # and round to it as accuracy
1334 $scale = $params[0]+4; # at least four more for proper round
1335 $params[2] = $r; # round mode by caller or undef
1336 $fallback = 1; # to clear a/p afterwards
1340 # the 4 below is empirical, and there might be cases where it is not
1342 $scale = abs($params[0] || $params[1]) + 4; # take whatever is defined
1345 my $rem; $rem = $self->bzero() if wantarray;
1347 $y = $self->new($y) unless $y->isa('Math::BigFloat');
1349 my $lx = $MBI->_len($x->{_m}); my $ly = $MBI->_len($y->{_m});
1350 $scale = $lx if $lx > $scale;
1351 $scale = $ly if $ly > $scale;
1352 my $diff = $ly - $lx;
1353 $scale += $diff if $diff > 0; # if lx << ly, but not if ly << lx!
1355 # already handled inf/NaN/-inf above:
1357 # check that $y is not 1 nor -1 and cache the result:
1358 my $y_not_one = !($MBI->_is_zero($y->{_e}) && $MBI->_is_one($y->{_m}));
1360 # flipping the sign of $y will also flip the sign of $x for the special
1361 # case of $x->bsub($x); so we can catch it below:
1362 my $xsign = $x->{sign};
1363 $y->{sign} =~ tr/+-/-+/;
1365 if ($xsign ne $x->{sign})
1367 # special case of $x /= $x results in 1
1368 $x->bone(); # "fixes" also sign of $y, since $x is $y
1372 # correct $y's sign again
1373 $y->{sign} =~ tr/+-/-+/;
1374 # continue with normal div code:
1376 # make copy of $x in case of list context for later reminder calculation
1377 if (wantarray && $y_not_one)
1382 $x->{sign} = $x->{sign} ne $y->sign() ? '-' : '+';
1384 # check for / +-1 ( +/- 1E0)
1387 # promote BigInts and it's subclasses (except when already a BigFloat)
1388 $y = $self->new($y) unless $y->isa('Math::BigFloat');
1390 # calculate the result to $scale digits and then round it
1391 # a * 10 ** b / c * 10 ** d => a/c * 10 ** (b-d)
1392 $MBI->_lsft($x->{_m},$MBI->_new($scale),10);
1393 $MBI->_div ($x->{_m},$y->{_m}); # a/c
1395 # correct exponent of $x
1396 ($x->{_e},$x->{_es}) = _e_sub($x->{_e}, $y->{_e}, $x->{_es}, $y->{_es});
1397 # correct for 10**scale
1398 ($x->{_e},$x->{_es}) = _e_sub($x->{_e}, $MBI->_new($scale), $x->{_es}, '+');
1399 $x->bnorm(); # remove trailing 0's
1401 } # ende else $x != $y
1403 # shortcut to not run through _find_round_parameters again
1404 if (defined $params[0])
1406 delete $x->{_a}; # clear before round
1407 $x->bround($params[0],$params[2]); # then round accordingly
1411 delete $x->{_p}; # clear before round
1412 $x->bfround($params[1],$params[2]); # then round accordingly
1416 # clear a/p after round, since user did not request it
1417 delete $x->{_a}; delete $x->{_p};
1424 $rem->bmod($y,@params); # copy already done
1428 # clear a/p after round, since user did not request it
1429 delete $rem->{_a}; delete $rem->{_p};
1438 # (dividend: BFLOAT or num_str, divisor: BFLOAT or num_str) return reminder
1441 my ($self,$x,$y,$a,$p,$r) = (ref($_[0]),@_);
1442 # objectify is costly, so avoid it
1443 if ((!ref($_[0])) || (ref($_[0]) ne ref($_[1])))
1445 ($self,$x,$y,$a,$p,$r) = objectify(2,@_);
1448 # handle NaN, inf, -inf
1449 if (($x->{sign} !~ /^[+-]$/) || ($y->{sign} !~ /^[+-]$/))
1451 my ($d,$re) = $self->SUPER::_div_inf($x,$y);
1452 $x->{sign} = $re->{sign};
1453 $x->{_e} = $re->{_e};
1454 $x->{_m} = $re->{_m};
1455 return $x->round($a,$p,$r,$y);
1459 return $x->bnan() if $x->is_zero();
1463 return $x->bzero() if $x->is_zero()
1465 # check that $y == +1 or $y == -1:
1466 ($MBI->_is_zero($y->{_e}) && $MBI->_is_one($y->{_m})));
1468 my $cmp = $x->bacmp($y); # equal or $x < $y?
1469 return $x->bzero($a,$p) if $cmp == 0; # $x == $y => result 0
1471 # only $y of the operands negative?
1472 my $neg = 0; $neg = 1 if $x->{sign} ne $y->{sign};
1474 $x->{sign} = $y->{sign}; # calc sign first
1475 return $x->round($a,$p,$r) if $cmp < 0 && $neg == 0; # $x < $y => result $x
1477 my $ym = $MBI->_copy($y->{_m});
1480 $MBI->_lsft( $ym, $y->{_e}, 10)
1481 if $y->{_es} eq '+' && !$MBI->_is_zero($y->{_e});
1483 # if $y has digits after dot
1484 my $shifty = 0; # correct _e of $x by this
1485 if ($y->{_es} eq '-') # has digits after dot
1487 # 123 % 2.5 => 1230 % 25 => 5 => 0.5
1488 $shifty = $MBI->_num($y->{_e}); # no more digits after dot
1489 $MBI->_lsft($x->{_m}, $y->{_e}, 10);# 123 => 1230, $y->{_m} is already 25
1491 # $ym is now mantissa of $y based on exponent 0
1493 my $shiftx = 0; # correct _e of $x by this
1494 if ($x->{_es} eq '-') # has digits after dot
1496 # 123.4 % 20 => 1234 % 200
1497 $shiftx = $MBI->_num($x->{_e}); # no more digits after dot
1498 $MBI->_lsft($ym, $x->{_e}, 10); # 123 => 1230
1500 # 123e1 % 20 => 1230 % 20
1501 if ($x->{_es} eq '+' && !$MBI->_is_zero($x->{_e}))
1503 $MBI->_lsft( $x->{_m}, $x->{_e},10); # es => '+' here
1506 $x->{_e} = $MBI->_new($shiftx);
1508 $x->{_es} = '-' if $shiftx != 0 || $shifty != 0;
1509 $MBI->_add( $x->{_e}, $MBI->_new($shifty)) if $shifty != 0;
1511 # now mantissas are equalized, exponent of $x is adjusted, so calc result
1513 $x->{_m} = $MBI->_mod( $x->{_m}, $ym);
1515 $x->{sign} = '+' if $MBI->_is_zero($x->{_m}); # fix sign for -0
1518 if ($neg != 0) # one of them negative => correct in place
1521 $x->{_m} = $r->{_m};
1522 $x->{_e} = $r->{_e};
1523 $x->{_es} = $r->{_es};
1524 $x->{sign} = '+' if $MBI->_is_zero($x->{_m}); # fix sign for -0
1528 $x->round($a,$p,$r,$y); # round and return
1533 # calculate $y'th root of $x
1536 my ($self,$x,$y,$a,$p,$r) = (ref($_[0]),@_);
1537 # objectify is costly, so avoid it
1538 if ((!ref($_[0])) || (ref($_[0]) ne ref($_[1])))
1540 ($self,$x,$y,$a,$p,$r) = objectify(2,@_);
1543 # NaN handling: $x ** 1/0, x or y NaN, or y inf/-inf or y == 0
1544 return $x->bnan() if $x->{sign} !~ /^\+/ || $y->is_zero() ||
1545 $y->{sign} !~ /^\+$/;
1547 return $x if $x->is_zero() || $x->is_one() || $x->is_inf() || $y->is_one();
1549 # we need to limit the accuracy to protect against overflow
1551 my (@params,$scale);
1552 ($x,@params) = $x->_find_round_parameters($a,$p,$r);
1554 return $x if $x->is_nan(); # error in _find_round_parameters?
1556 # no rounding at all, so must use fallback
1557 if (scalar @params == 0)
1559 # simulate old behaviour
1560 $params[0] = $self->div_scale(); # and round to it as accuracy
1561 $scale = $params[0]+4; # at least four more for proper round
1562 $params[2] = $r; # iound mode by caller or undef
1563 $fallback = 1; # to clear a/p afterwards
1567 # the 4 below is empirical, and there might be cases where it is not
1569 $scale = abs($params[0] || $params[1]) + 4; # take whatever is defined
1572 # when user set globals, they would interfere with our calculation, so
1573 # disable them and later re-enable them
1575 my $abr = "$self\::accuracy"; my $ab = $$abr; $$abr = undef;
1576 my $pbr = "$self\::precision"; my $pb = $$pbr; $$pbr = undef;
1577 # we also need to disable any set A or P on $x (_find_round_parameters took
1578 # them already into account), since these would interfere, too
1579 delete $x->{_a}; delete $x->{_p};
1580 # need to disable $upgrade in BigInt, to avoid deep recursion
1581 local $Math::BigInt::upgrade = undef; # should be really parent class vs MBI
1583 # remember sign and make $x positive, since -4 ** (1/2) => -2
1584 my $sign = 0; $sign = 1 if $x->{sign} eq '-'; $x->{sign} = '+';
1587 if ($y->isa('Math::BigFloat'))
1589 $is_two = ($y->{sign} eq '+' && $MBI->_is_two($y->{_m}) && $MBI->_is_zero($y->{_e}));
1593 $is_two = ($y == 2);
1596 # normal square root if $y == 2:
1599 $x->bsqrt($scale+4);
1601 elsif ($y->is_one('-'))
1604 my $u = $self->bone()->bdiv($x,$scale);
1605 # copy private parts over
1606 $x->{_m} = $u->{_m};
1607 $x->{_e} = $u->{_e};
1608 $x->{_es} = $u->{_es};
1612 # calculate the broot() as integer result first, and if it fits, return
1613 # it rightaway (but only if $x and $y are integer):
1615 my $done = 0; # not yet
1616 if ($y->is_int() && $x->is_int())
1618 my $i = $MBI->_copy( $x->{_m} );
1619 $MBI->_lsft( $i, $x->{_e}, 10 ) unless $MBI->_is_zero($x->{_e});
1620 my $int = Math::BigInt->bzero();
1622 $int->broot($y->as_number());
1624 if ($int->copy()->bpow($y) == $x)
1626 # found result, return it
1627 $x->{_m} = $int->{value};
1628 $x->{_e} = $MBI->_zero();
1636 my $u = $self->bone()->bdiv($y,$scale+4);
1637 delete $u->{_a}; delete $u->{_p}; # otherwise it conflicts
1638 $x->bpow($u,$scale+4); # el cheapo
1641 $x->bneg() if $sign == 1;
1643 # shortcut to not run through _find_round_parameters again
1644 if (defined $params[0])
1646 $x->bround($params[0],$params[2]); # then round accordingly
1650 $x->bfround($params[1],$params[2]); # then round accordingly
1654 # clear a/p after round, since user did not request it
1655 delete $x->{_a}; delete $x->{_p};
1658 $$abr = $ab; $$pbr = $pb;
1664 # calculate square root
1665 my ($self,$x,$a,$p,$r) = ref($_[0]) ? (ref($_[0]),@_) : objectify(1,@_);
1667 return $x->bnan() if $x->{sign} !~ /^[+]/; # NaN, -inf or < 0
1668 return $x if $x->{sign} eq '+inf'; # sqrt(inf) == inf
1669 return $x->round($a,$p,$r) if $x->is_zero() || $x->is_one();
1671 # we need to limit the accuracy to protect against overflow
1673 my (@params,$scale);
1674 ($x,@params) = $x->_find_round_parameters($a,$p,$r);
1676 return $x if $x->is_nan(); # error in _find_round_parameters?
1678 # no rounding at all, so must use fallback
1679 if (scalar @params == 0)
1681 # simulate old behaviour
1682 $params[0] = $self->div_scale(); # and round to it as accuracy
1683 $scale = $params[0]+4; # at least four more for proper round
1684 $params[2] = $r; # round mode by caller or undef
1685 $fallback = 1; # to clear a/p afterwards
1689 # the 4 below is empirical, and there might be cases where it is not
1691 $scale = abs($params[0] || $params[1]) + 4; # take whatever is defined
1694 # when user set globals, they would interfere with our calculation, so
1695 # disable them and later re-enable them
1697 my $abr = "$self\::accuracy"; my $ab = $$abr; $$abr = undef;
1698 my $pbr = "$self\::precision"; my $pb = $$pbr; $$pbr = undef;
1699 # we also need to disable any set A or P on $x (_find_round_parameters took
1700 # them already into account), since these would interfere, too
1701 delete $x->{_a}; delete $x->{_p};
1702 # need to disable $upgrade in BigInt, to avoid deep recursion
1703 local $Math::BigInt::upgrade = undef; # should be really parent class vs MBI
1705 my $i = $MBI->_copy( $x->{_m} );
1706 $MBI->_lsft( $i, $x->{_e}, 10 ) unless $MBI->_is_zero($x->{_e});
1707 my $xas = Math::BigInt->bzero();
1710 my $gs = $xas->copy()->bsqrt(); # some guess
1712 if (($x->{_es} ne '-') # guess can't be accurate if there are
1713 # digits after the dot
1714 && ($xas->bacmp($gs * $gs) == 0)) # guess hit the nail on the head?
1716 # exact result, copy result over to keep $x
1717 $x->{_m} = $gs->{value}; $x->{_e} = $MBI->_zero(); $x->{_es} = '+';
1719 # shortcut to not run through _find_round_parameters again
1720 if (defined $params[0])
1722 $x->bround($params[0],$params[2]); # then round accordingly
1726 $x->bfround($params[1],$params[2]); # then round accordingly
1730 # clear a/p after round, since user did not request it
1731 delete $x->{_a}; delete $x->{_p};
1733 # re-enable A and P, upgrade is taken care of by "local"
1734 ${"$self\::accuracy"} = $ab; ${"$self\::precision"} = $pb;
1738 # sqrt(2) = 1.4 because sqrt(2*100) = 1.4*10; so we can increase the accuracy
1739 # of the result by multipyling the input by 100 and then divide the integer
1740 # result of sqrt(input) by 10. Rounding afterwards returns the real result.
1742 # The following steps will transform 123.456 (in $x) into 123456 (in $y1)
1743 my $y1 = $MBI->_copy($x->{_m});
1745 my $length = $MBI->_len($y1);
1747 # Now calculate how many digits the result of sqrt(y1) would have
1748 my $digits = int($length / 2);
1750 # But we need at least $scale digits, so calculate how many are missing
1751 my $shift = $scale - $digits;
1753 # That should never happen (we take care of integer guesses above)
1754 # $shift = 0 if $shift < 0;
1756 # Multiply in steps of 100, by shifting left two times the "missing" digits
1757 my $s2 = $shift * 2;
1759 # We now make sure that $y1 has the same odd or even number of digits than
1760 # $x had. So when _e of $x is odd, we must shift $y1 by one digit left,
1761 # because we always must multiply by steps of 100 (sqrt(100) is 10) and not
1762 # steps of 10. The length of $x does not count, since an even or odd number
1763 # of digits before the dot is not changed by adding an even number of digits
1764 # after the dot (the result is still odd or even digits long).
1765 $s2++ if $MBI->_is_odd($x->{_e});
1767 $MBI->_lsft( $y1, $MBI->_new($s2), 10);
1769 # now take the square root and truncate to integer
1770 $y1 = $MBI->_sqrt($y1);
1772 # By "shifting" $y1 right (by creating a negative _e) we calculate the final
1773 # result, which is than later rounded to the desired scale.
1775 # calculate how many zeros $x had after the '.' (or before it, depending
1776 # on sign of $dat, the result should have half as many:
1777 my $dat = $MBI->_num($x->{_e});
1778 $dat = -$dat if $x->{_es} eq '-';
1783 # no zeros after the dot (e.g. 1.23, 0.49 etc)
1784 # preserve half as many digits before the dot than the input had
1785 # (but round this "up")
1786 $dat = int(($dat+1)/2);
1790 $dat = int(($dat)/2);
1792 $dat -= $MBI->_len($y1);
1796 $x->{_e} = $MBI->_new( $dat );
1801 $x->{_e} = $MBI->_new( $dat );
1807 # shortcut to not run through _find_round_parameters again
1808 if (defined $params[0])
1810 $x->bround($params[0],$params[2]); # then round accordingly
1814 $x->bfround($params[1],$params[2]); # then round accordingly
1818 # clear a/p after round, since user did not request it
1819 delete $x->{_a}; delete $x->{_p};
1822 $$abr = $ab; $$pbr = $pb;
1828 # (BFLOAT or num_str, BFLOAT or num_str) return BFLOAT
1829 # compute factorial number, modifies first argument
1832 my ($self,$x,@r) = (ref($_[0]),@_);
1833 # objectify is costly, so avoid it
1834 ($self,$x,@r) = objectify(1,@_) if !ref($x);
1836 return $x if $x->{sign} eq '+inf'; # inf => inf
1838 if (($x->{sign} ne '+') || # inf, NaN, <0 etc => NaN
1839 ($x->{_es} ne '+')); # digits after dot?
1841 # use BigInt's bfac() for faster calc
1842 if (! $MBI->_is_zero($x->{_e}))
1844 $MBI->_lsft($x->{_m}, $x->{_e},10); # change 12e1 to 120e0
1845 $x->{_e} = $MBI->_zero(); # normalize
1848 $MBI->_fac($x->{_m}); # calculate factorial
1849 $x->bnorm()->round(@r); # norm again and round result
1854 # Calculate a power where $y is a non-integer, like 2 ** 0.5
1855 my ($x,$y,$a,$p,$r) = @_;
1858 # if $y == 0.5, it is sqrt($x)
1859 $HALF = $self->new($HALF) unless ref($HALF);
1860 return $x->bsqrt($a,$p,$r,$y) if $y->bcmp($HALF) == 0;
1863 # a ** x == e ** (x * ln a)
1867 # Taylor: | u u^2 u^3 |
1868 # x ** y = 1 + | --- + --- + ----- + ... |
1871 # we need to limit the accuracy to protect against overflow
1873 my ($scale,@params);
1874 ($x,@params) = $x->_find_round_parameters($a,$p,$r);
1876 return $x if $x->is_nan(); # error in _find_round_parameters?
1878 # no rounding at all, so must use fallback
1879 if (scalar @params == 0)
1881 # simulate old behaviour
1882 $params[0] = $self->div_scale(); # and round to it as accuracy
1883 $params[1] = undef; # disable P
1884 $scale = $params[0]+4; # at least four more for proper round
1885 $params[2] = $r; # round mode by caller or undef
1886 $fallback = 1; # to clear a/p afterwards
1890 # the 4 below is empirical, and there might be cases where it is not
1892 $scale = abs($params[0] || $params[1]) + 4; # take whatever is defined
1895 # when user set globals, they would interfere with our calculation, so
1896 # disable them and later re-enable them
1898 my $abr = "$self\::accuracy"; my $ab = $$abr; $$abr = undef;
1899 my $pbr = "$self\::precision"; my $pb = $$pbr; $$pbr = undef;
1900 # we also need to disable any set A or P on $x (_find_round_parameters took
1901 # them already into account), since these would interfere, too
1902 delete $x->{_a}; delete $x->{_p};
1903 # need to disable $upgrade in BigInt, to avoid deep recursion
1904 local $Math::BigInt::upgrade = undef;
1906 my ($limit,$v,$u,$below,$factor,$next,$over);
1908 $u = $x->copy()->blog(undef,$scale)->bmul($y);
1909 $v = $self->bone(); # 1
1910 $factor = $self->new(2); # 2
1911 $x->bone(); # first term: 1
1913 $below = $v->copy();
1916 $limit = $self->new("1E-". ($scale-1));
1920 # we calculate the next term, and add it to the last
1921 # when the next term is below our limit, it won't affect the outcome
1922 # anymore, so we stop
1923 $next = $over->copy()->bdiv($below,$scale);
1924 last if $next->bacmp($limit) <= 0;
1926 # calculate things for the next term
1927 $over *= $u; $below *= $factor; $factor->binc();
1929 last if $x->{sign} !~ /^[-+]$/;
1934 # shortcut to not run through _find_round_parameters again
1935 if (defined $params[0])
1937 $x->bround($params[0],$params[2]); # then round accordingly
1941 $x->bfround($params[1],$params[2]); # then round accordingly
1945 # clear a/p after round, since user did not request it
1946 delete $x->{_a}; delete $x->{_p};
1949 $$abr = $ab; $$pbr = $pb;
1955 # (BFLOAT or num_str, BFLOAT or num_str) return BFLOAT
1956 # compute power of two numbers, second arg is used as integer
1957 # modifies first argument
1960 my ($self,$x,$y,$a,$p,$r) = (ref($_[0]),@_);
1961 # objectify is costly, so avoid it
1962 if ((!ref($_[0])) || (ref($_[0]) ne ref($_[1])))
1964 ($self,$x,$y,$a,$p,$r) = objectify(2,@_);
1967 return $x->bnan() if $x->{sign} eq $nan || $y->{sign} eq $nan;
1968 return $x if $x->{sign} =~ /^[+-]inf$/;
1970 # cache the result of is_zero
1971 my $y_is_zero = $y->is_zero();
1972 return $x->bone() if $y_is_zero;
1973 return $x if $x->is_one() || $y->is_one();
1975 my $x_is_zero = $x->is_zero();
1976 return $x->_pow($y,$a,$p,$r) if !$x_is_zero && !$y->is_int(); # non-integer power
1978 my $y1 = $y->as_number()->{value}; # make MBI part
1981 if ($x->{sign} eq '-' && $MBI->_is_one($x->{_m}) && $MBI->_is_zero($x->{_e}))
1983 # if $x == -1 and odd/even y => +1/-1 because +-1 ^ (+-1) => +-1
1984 return $MBI->_is_odd($y1) ? $x : $x->babs(1);
1988 return $x->bone() if $y_is_zero;
1989 return $x if $y->{sign} eq '+'; # 0**y => 0 (if not y <= 0)
1990 # 0 ** -y => 1 / (0 ** y) => 1 / 0! (1 / 0 => +inf)
1995 $new_sign = $MBI->_is_odd($y1) ? '-' : '+' if $x->{sign} ne '+';
1997 # calculate $x->{_m} ** $y and $x->{_e} * $y separately (faster)
1998 $x->{_m} = $MBI->_pow( $x->{_m}, $y1);
1999 $x->{_e} = $MBI->_mul ($x->{_e}, $y1);
2001 $x->{sign} = $new_sign;
2003 if ($y->{sign} eq '-')
2005 # modify $x in place!
2006 my $z = $x->copy(); $x->bone();
2007 return scalar $x->bdiv($z,$a,$p,$r); # round in one go (might ignore y's A!)
2009 $x->round($a,$p,$r,$y);
2012 ###############################################################################
2013 # rounding functions
2017 # precision: round to the $Nth digit left (+$n) or right (-$n) from the '.'
2018 # $n == 0 means round to integer
2019 # expects and returns normalized numbers!
2020 my $x = shift; my $self = ref($x) || $x; $x = $self->new(shift) if !ref($x);
2022 my ($scale,$mode) = $x->_scale_p(@_);
2023 return $x if !defined $scale || $x->modify('bfround'); # no-op
2025 # never round a 0, +-inf, NaN
2028 $x->{_p} = $scale if !defined $x->{_p} || $x->{_p} < $scale; # -3 < -2
2031 return $x if $x->{sign} !~ /^[+-]$/;
2033 # don't round if x already has lower precision
2034 return $x if (defined $x->{_p} && $x->{_p} < 0 && $scale < $x->{_p});
2036 $x->{_p} = $scale; # remember round in any case
2037 delete $x->{_a}; # and clear A
2040 # round right from the '.'
2042 return $x if $x->{_es} eq '+'; # e >= 0 => nothing to round
2044 $scale = -$scale; # positive for simplicity
2045 my $len = $MBI->_len($x->{_m}); # length of mantissa
2047 # the following poses a restriction on _e, but if _e is bigger than a
2048 # scalar, you got other problems (memory etc) anyway
2049 my $dad = -(0+ ($x->{_es}.$MBI->_num($x->{_e}))); # digits after dot
2050 my $zad = 0; # zeros after dot
2051 $zad = $dad - $len if (-$dad < -$len); # for 0.00..00xxx style
2053 # p rint "scale $scale dad $dad zad $zad len $len\n";
2054 # number bsstr len zad dad
2055 # 0.123 123e-3 3 0 3
2056 # 0.0123 123e-4 3 1 4
2059 # 1.2345 12345e-4 5 0 4
2061 # do not round after/right of the $dad
2062 return $x if $scale > $dad; # 0.123, scale >= 3 => exit
2064 # round to zero if rounding inside the $zad, but not for last zero like:
2065 # 0.0065, scale -2, round last '0' with following '65' (scale == zad case)
2066 return $x->bzero() if $scale < $zad;
2067 if ($scale == $zad) # for 0.006, scale -3 and trunc
2073 # adjust round-point to be inside mantissa
2076 $scale = $scale-$zad;
2080 my $dbd = $len - $dad; $dbd = 0 if $dbd < 0; # digits before dot
2081 $scale = $dbd+$scale;
2087 # round left from the '.'
2089 # 123 => 100 means length(123) = 3 - $scale (2) => 1
2091 my $dbt = $MBI->_len($x->{_m});
2093 my $dbd = $dbt + ($x->{_es} . $MBI->_num($x->{_e}));
2094 # should be the same, so treat it as this
2095 $scale = 1 if $scale == 0;
2096 # shortcut if already integer
2097 return $x if $scale == 1 && $dbt <= $dbd;
2098 # maximum digits before dot
2103 # not enough digits before dot, so round to zero
2106 elsif ( $scale == $dbd )
2113 $scale = $dbd - $scale;
2116 # pass sign to bround for rounding modes '+inf' and '-inf'
2117 my $m = bless { sign => $x->{sign}, value => $x->{_m} }, 'Math::BigInt';
2118 $m->bround($scale,$mode);
2119 $x->{_m} = $m->{value}; # get our mantissa back
2125 # accuracy: preserve $N digits, and overwrite the rest with 0's
2126 my $x = shift; my $self = ref($x) || $x; $x = $self->new(shift) if !ref($x);
2128 if (($_[0] || 0) < 0)
2130 require Carp; Carp::croak ('bround() needs positive accuracy');
2133 my ($scale,$mode) = $x->_scale_a(@_);
2134 return $x if !defined $scale || $x->modify('bround'); # no-op
2136 # scale is now either $x->{_a}, $accuracy, or the user parameter
2137 # test whether $x already has lower accuracy, do nothing in this case
2138 # but do round if the accuracy is the same, since a math operation might
2139 # want to round a number with A=5 to 5 digits afterwards again
2140 return $x if defined $x->{_a} && $x->{_a} < $scale;
2142 # scale < 0 makes no sense
2143 # scale == 0 => keep all digits
2144 # never round a +-inf, NaN
2145 return $x if ($scale <= 0) || $x->{sign} !~ /^[+-]$/;
2147 # 1: never round a 0
2148 # 2: if we should keep more digits than the mantissa has, do nothing
2149 if ($x->is_zero() || $MBI->_len($x->{_m}) <= $scale)
2151 $x->{_a} = $scale if !defined $x->{_a} || $x->{_a} > $scale;
2155 # pass sign to bround for '+inf' and '-inf' rounding modes
2156 my $m = bless { sign => $x->{sign}, value => $x->{_m} }, 'Math::BigInt';
2158 $m->bround($scale,$mode); # round mantissa
2159 $x->{_m} = $m->{value}; # get our mantissa back
2160 $x->{_a} = $scale; # remember rounding
2161 delete $x->{_p}; # and clear P
2162 $x->bnorm(); # del trailing zeros gen. by bround()
2167 # return integer less or equal then $x
2168 my ($self,$x,$a,$p,$r) = ref($_[0]) ? (ref($_[0]),@_) : objectify(1,@_);
2170 return $x if $x->modify('bfloor');
2172 return $x if $x->{sign} !~ /^[+-]$/; # nan, +inf, -inf
2174 # if $x has digits after dot
2175 if ($x->{_es} eq '-')
2177 $x->{_m} = $MBI->_rsft($x->{_m},$x->{_e},10); # cut off digits after dot
2178 $x->{_e} = $MBI->_zero(); # trunc/norm
2179 $x->{_es} = '+'; # abs e
2180 $MBI->_inc($x->{_m}) if $x->{sign} eq '-'; # increment if negative
2182 $x->round($a,$p,$r);
2187 # return integer greater or equal then $x
2188 my ($self,$x,$a,$p,$r) = ref($_[0]) ? (ref($_[0]),@_) : objectify(1,@_);
2190 return $x if $x->modify('bceil');
2191 return $x if $x->{sign} !~ /^[+-]$/; # nan, +inf, -inf
2193 # if $x has digits after dot
2194 if ($x->{_es} eq '-')
2196 $x->{_m} = $MBI->_rsft($x->{_m},$x->{_e},10); # cut off digits after dot
2197 $x->{_e} = $MBI->_zero(); # trunc/norm
2198 $x->{_es} = '+'; # abs e
2199 $MBI->_inc($x->{_m}) if $x->{sign} eq '+'; # increment if positive
2201 $x->round($a,$p,$r);
2206 # shift right by $y (divide by power of $n)
2209 my ($self,$x,$y,$n,$a,$p,$r) = (ref($_[0]),@_);
2210 # objectify is costly, so avoid it
2211 if ((!ref($_[0])) || (ref($_[0]) ne ref($_[1])))
2213 ($self,$x,$y,$n,$a,$p,$r) = objectify(2,@_);
2216 return $x if $x->modify('brsft');
2217 return $x if $x->{sign} !~ /^[+-]$/; # nan, +inf, -inf
2219 $n = 2 if !defined $n; $n = $self->new($n);
2222 return $x->blsft($y->copy()->babs(),$n) if $y->{sign} =~ /^-/;
2224 # the following call to bdiv() will return either quo or (quo,reminder):
2225 $x->bdiv($n->bpow($y),$a,$p,$r,$y);
2230 # shift left by $y (multiply by power of $n)
2233 my ($self,$x,$y,$n,$a,$p,$r) = (ref($_[0]),@_);
2234 # objectify is costly, so avoid it
2235 if ((!ref($_[0])) || (ref($_[0]) ne ref($_[1])))
2237 ($self,$x,$y,$n,$a,$p,$r) = objectify(2,@_);
2240 return $x if $x->modify('blsft');
2241 return $x if $x->{sign} !~ /^[+-]$/; # nan, +inf, -inf
2243 $n = 2 if !defined $n; $n = $self->new($n);
2246 return $x->brsft($y->copy()->babs(),$n) if $y->{sign} =~ /^-/;
2248 $x->bmul($n->bpow($y),$a,$p,$r,$y);
2251 ###############################################################################
2255 # going through AUTOLOAD for every DESTROY is costly, avoid it by empty sub
2260 # make fxxx and bxxx both work by selectively mapping fxxx() to MBF::bxxx()
2261 # or falling back to MBI::bxxx()
2262 my $name = $AUTOLOAD;
2264 $name =~ s/(.*):://; # split package
2265 my $c = $1 || $class;
2267 $c->import() if $IMPORT == 0;
2268 if (!_method_alias($name))
2272 # delayed load of Carp and avoid recursion
2274 Carp::croak ("$c: Can't call a method without name");
2276 if (!_method_hand_up($name))
2278 # delayed load of Carp and avoid recursion
2280 Carp::croak ("Can't call $c\-\>$name, not a valid method");
2282 # try one level up, but subst. bxxx() for fxxx() since MBI only got bxxx()
2284 return &{"Math::BigInt"."::$name"}(@_);
2286 my $bname = $name; $bname =~ s/^f/b/;
2294 # return a copy of the exponent
2295 my ($self,$x) = ref($_[0]) ? (ref($_[0]),$_[0]) : objectify(1,@_);
2297 if ($x->{sign} !~ /^[+-]$/)
2299 my $s = $x->{sign}; $s =~ s/^[+-]//;
2300 return Math::BigInt->new($s); # -inf, +inf => +inf
2302 Math::BigInt->new( $x->{_es} . $MBI->_str($x->{_e}));
2307 # return a copy of the mantissa
2308 my ($self,$x) = ref($_[0]) ? (ref($_[0]),$_[0]) : objectify(1,@_);
2310 if ($x->{sign} !~ /^[+-]$/)
2312 my $s = $x->{sign}; $s =~ s/^[+]//;
2313 return Math::BigInt->new($s); # -inf, +inf => +inf
2315 my $m = Math::BigInt->new( $MBI->_str($x->{_m}));
2316 $m->bneg() if $x->{sign} eq '-';
2323 # return a copy of both the exponent and the mantissa
2324 my ($self,$x) = ref($_[0]) ? (ref($_[0]),$_[0]) : objectify(1,@_);
2326 if ($x->{sign} !~ /^[+-]$/)
2328 my $s = $x->{sign}; $s =~ s/^[+]//; my $se = $s; $se =~ s/^[-]//;
2329 return ($self->new($s),$self->new($se)); # +inf => inf and -inf,+inf => inf
2331 my $m = Math::BigInt->bzero();
2332 $m->{value} = $MBI->_copy($x->{_m});
2333 $m->bneg() if $x->{sign} eq '-';
2334 ($m, Math::BigInt->new( $x->{_es} . $MBI->_num($x->{_e}) ));
2337 ##############################################################################
2338 # private stuff (internal use only)
2344 my $lib = ''; my @a;
2346 for ( my $i = 0; $i < $l ; $i++)
2348 if ( $_[$i] eq ':constant' )
2350 # This causes overlord er load to step in. 'binary' and 'integer'
2351 # are handled by BigInt.
2352 overload::constant float => sub { $self->new(shift); };
2354 elsif ($_[$i] eq 'upgrade')
2356 # this causes upgrading
2357 $upgrade = $_[$i+1]; # or undef to disable
2360 elsif ($_[$i] eq 'downgrade')
2362 # this causes downgrading
2363 $downgrade = $_[$i+1]; # or undef to disable
2366 elsif ($_[$i] eq 'lib')
2368 # alternative library
2369 $lib = $_[$i+1] || ''; # default Calc
2372 elsif ($_[$i] eq 'with')
2374 # alternative class for our private parts()
2375 # XXX: no longer supported
2376 # $MBI = $_[$i+1] || 'Math::BigInt';
2385 $lib =~ tr/a-zA-Z0-9,://cd; # restrict to sane characters
2386 # let use Math::BigInt lib => 'GMP'; use Math::BigFloat; still work
2387 my $mbilib = eval { Math::BigInt->config()->{lib} };
2388 if ((defined $mbilib) && ($MBI eq 'Math::BigInt::Calc'))
2390 # MBI already loaded
2391 Math::BigInt->import('try',"$lib,$mbilib", 'objectify');
2395 # MBI not loaded, or with ne "Math::BigInt::Calc"
2396 $lib .= ",$mbilib" if defined $mbilib;
2397 $lib =~ s/^,//; # don't leave empty
2399 # replacement library can handle lib statement, but also could ignore it
2401 # Perl < 5.6.0 dies with "out of memory!" when eval() and ':constant' is
2402 # used in the same script, or eval inside import(). So we require MBI:
2403 require Math::BigInt;
2404 Math::BigInt->import( try => $lib, 'objectify' );
2408 require Carp; Carp::croak ("Couldn't load $lib: $! $@");
2410 # find out which one was actually loaded
2411 $MBI = Math::BigInt->config()->{lib};
2413 # register us with MBI to get notified of future lib changes
2414 Math::BigInt::_register_callback( $self, sub { $MBI = $_[0]; } );
2416 # any non :constant stuff is handled by our parent, Exporter
2417 # even if @_ is empty, to give it a chance
2418 $self->SUPER::import(@a); # for subclasses
2419 $self->export_to_level(1,$self,@a); # need this, too
2424 # adjust m and e so that m is smallest possible
2425 # round number according to accuracy and precision settings
2426 my ($self,$x) = ref($_[0]) ? (undef,$_[0]) : objectify(1,@_);
2428 return $x if $x->{sign} !~ /^[+-]$/; # inf, nan etc
2430 my $zeros = $MBI->_zeros($x->{_m}); # correct for trailing zeros
2433 my $z = $MBI->_new($zeros);
2434 $x->{_m} = $MBI->_rsft ($x->{_m}, $z, 10);
2435 if ($x->{_es} eq '-')
2437 if ($MBI->_acmp($x->{_e},$z) >= 0)
2439 $x->{_e} = $MBI->_sub ($x->{_e}, $z);
2440 $x->{_es} = '+' if $MBI->_is_zero($x->{_e});
2444 $x->{_e} = $MBI->_sub ( $MBI->_copy($z), $x->{_e});
2450 $x->{_e} = $MBI->_add ($x->{_e}, $z);
2455 # $x can only be 0Ey if there are no trailing zeros ('0' has 0 trailing
2456 # zeros). So, for something like 0Ey, set y to 1, and -0 => +0
2457 $x->{sign} = '+', $x->{_es} = '+', $x->{_e} = $MBI->_one()
2458 if $MBI->_is_zero($x->{_m});
2461 $x; # MBI bnorm is no-op, so dont call it
2464 ##############################################################################
2468 # return number as hexadecimal string (only for integers defined)
2469 my ($self,$x) = ref($_[0]) ? (ref($_[0]),$_[0]) : objectify(1,@_);
2471 return $x->bstr() if $x->{sign} !~ /^[+-]$/; # inf, nan etc
2472 return '0x0' if $x->is_zero();
2474 return $nan if $x->{_es} ne '+'; # how to do 1e-1 in hex!?
2476 my $z = $MBI->_copy($x->{_m});
2477 if (! $MBI->_is_zero($x->{_e})) # > 0
2479 $MBI->_lsft( $z, $x->{_e},10);
2481 $z = Math::BigInt->new( $x->{sign} . $MBI->_num($z));
2487 # return number as binary digit string (only for integers defined)
2488 my ($self,$x) = ref($_[0]) ? (ref($_[0]),$_[0]) : objectify(1,@_);
2490 return $x->bstr() if $x->{sign} !~ /^[+-]$/; # inf, nan etc
2491 return '0b0' if $x->is_zero();
2493 return $nan if $x->{_es} ne '+'; # how to do 1e-1 in hex!?
2495 my $z = $MBI->_copy($x->{_m});
2496 if (! $MBI->_is_zero($x->{_e})) # > 0
2498 $MBI->_lsft( $z, $x->{_e},10);
2500 $z = Math::BigInt->new( $x->{sign} . $MBI->_num($z));
2506 # return number as octal digit string (only for integers defined)
2507 my ($self,$x) = ref($_[0]) ? (ref($_[0]),$_[0]) : objectify(1,@_);
2509 return $x->bstr() if $x->{sign} !~ /^[+-]$/; # inf, nan etc
2510 return '0' if $x->is_zero();
2512 return $nan if $x->{_es} ne '+'; # how to do 1e-1 in hex!?
2514 my $z = $MBI->_copy($x->{_m});
2515 if (! $MBI->_is_zero($x->{_e})) # > 0
2517 $MBI->_lsft( $z, $x->{_e},10);
2519 $z = Math::BigInt->new( $x->{sign} . $MBI->_num($z));
2525 # return copy as a bigint representation of this BigFloat number
2526 my ($self,$x) = ref($_[0]) ? (ref($_[0]),$_[0]) : objectify(1,@_);
2528 my $z = $MBI->_copy($x->{_m});
2529 if ($x->{_es} eq '-') # < 0
2531 $MBI->_rsft( $z, $x->{_e},10);
2533 elsif (! $MBI->_is_zero($x->{_e})) # > 0
2535 $MBI->_lsft( $z, $x->{_e},10);
2537 $z = Math::BigInt->new( $x->{sign} . $MBI->_num($z));
2544 my $class = ref($x) || $x;
2545 $x = $class->new(shift) unless ref($x);
2547 return 1 if $MBI->_is_zero($x->{_m});
2549 my $len = $MBI->_len($x->{_m});
2550 $len += $MBI->_num($x->{_e}) if $x->{_es} eq '+';
2554 $t = $MBI->_num($x->{_e}) if $x->{_es} eq '-';
2565 Math::BigFloat - Arbitrary size floating point math package
2572 $x = Math::BigFloat->new($str); # defaults to 0
2573 $nan = Math::BigFloat->bnan(); # create a NotANumber
2574 $zero = Math::BigFloat->bzero(); # create a +0
2575 $inf = Math::BigFloat->binf(); # create a +inf
2576 $inf = Math::BigFloat->binf('-'); # create a -inf
2577 $one = Math::BigFloat->bone(); # create a +1
2578 $one = Math::BigFloat->bone('-'); # create a -1
2581 $x->is_zero(); # true if arg is +0
2582 $x->is_nan(); # true if arg is NaN
2583 $x->is_one(); # true if arg is +1
2584 $x->is_one('-'); # true if arg is -1
2585 $x->is_odd(); # true if odd, false for even
2586 $x->is_even(); # true if even, false for odd
2587 $x->is_pos(); # true if >= 0
2588 $x->is_neg(); # true if < 0
2589 $x->is_inf(sign); # true if +inf, or -inf (default is '+')
2591 $x->bcmp($y); # compare numbers (undef,<0,=0,>0)
2592 $x->bacmp($y); # compare absolutely (undef,<0,=0,>0)
2593 $x->sign(); # return the sign, either +,- or NaN
2594 $x->digit($n); # return the nth digit, counting from right
2595 $x->digit(-$n); # return the nth digit, counting from left
2597 # The following all modify their first argument. If you want to preserve
2598 # $x, use $z = $x->copy()->bXXX($y); See under L<CAVEATS> for why this is
2599 # necessary when mixing $a = $b assignments with non-overloaded math.
2602 $x->bzero(); # set $i to 0
2603 $x->bnan(); # set $i to NaN
2604 $x->bone(); # set $x to +1
2605 $x->bone('-'); # set $x to -1
2606 $x->binf(); # set $x to inf
2607 $x->binf('-'); # set $x to -inf
2609 $x->bneg(); # negation
2610 $x->babs(); # absolute value
2611 $x->bnorm(); # normalize (no-op)
2612 $x->bnot(); # two's complement (bit wise not)
2613 $x->binc(); # increment x by 1
2614 $x->bdec(); # decrement x by 1
2616 $x->badd($y); # addition (add $y to $x)
2617 $x->bsub($y); # subtraction (subtract $y from $x)
2618 $x->bmul($y); # multiplication (multiply $x by $y)
2619 $x->bdiv($y); # divide, set $x to quotient
2620 # return (quo,rem) or quo if scalar
2622 $x->bmod($y); # modulus ($x % $y)
2623 $x->bpow($y); # power of arguments ($x ** $y)
2624 $x->blsft($y, $n); # left shift by $y places in base $n
2625 $x->brsft($y, $n); # right shift by $y places in base $n
2626 # returns (quo,rem) or quo if in scalar context
2628 $x->blog(); # logarithm of $x to base e (Euler's number)
2629 $x->blog($base); # logarithm of $x to base $base (f.i. 2)
2631 $x->band($y); # bit-wise and
2632 $x->bior($y); # bit-wise inclusive or
2633 $x->bxor($y); # bit-wise exclusive or
2634 $x->bnot(); # bit-wise not (two's complement)
2636 $x->bsqrt(); # calculate square-root
2637 $x->broot($y); # $y'th root of $x (e.g. $y == 3 => cubic root)
2638 $x->bfac(); # factorial of $x (1*2*3*4*..$x)
2640 $x->bround($N); # accuracy: preserve $N digits
2641 $x->bfround($N); # precision: round to the $Nth digit
2643 $x->bfloor(); # return integer less or equal than $x
2644 $x->bceil(); # return integer greater or equal than $x
2646 # The following do not modify their arguments:
2648 bgcd(@values); # greatest common divisor
2649 blcm(@values); # lowest common multiplicator
2651 $x->bstr(); # return string
2652 $x->bsstr(); # return string in scientific notation
2654 $x->as_int(); # return $x as BigInt
2655 $x->exponent(); # return exponent as BigInt
2656 $x->mantissa(); # return mantissa as BigInt
2657 $x->parts(); # return (mantissa,exponent) as BigInt
2659 $x->length(); # number of digits (w/o sign and '.')
2660 ($l,$f) = $x->length(); # number of digits, and length of fraction
2662 $x->precision(); # return P of $x (or global, if P of $x undef)
2663 $x->precision($n); # set P of $x to $n
2664 $x->accuracy(); # return A of $x (or global, if A of $x undef)
2665 $x->accuracy($n); # set A $x to $n
2667 # these get/set the appropriate global value for all BigFloat objects
2668 Math::BigFloat->precision(); # Precision
2669 Math::BigFloat->accuracy(); # Accuracy
2670 Math::BigFloat->round_mode(); # rounding mode
2674 All operators (including basic math operations) are overloaded if you
2675 declare your big floating point numbers as
2677 $i = new Math::BigFloat '12_3.456_789_123_456_789E-2';
2679 Operations with overloaded operators preserve the arguments, which is
2680 exactly what you expect.
2682 =head2 Canonical notation
2684 Input to these routines are either BigFloat objects, or strings of the
2685 following four forms:
2699 C</^[+-]\d+E[+-]?\d+$/>
2703 C</^[+-]\d*\.\d+E[+-]?\d+$/>
2707 all with optional leading and trailing zeros and/or spaces. Additionally,
2708 numbers are allowed to have an underscore between any two digits.
2710 Empty strings as well as other illegal numbers results in 'NaN'.
2712 bnorm() on a BigFloat object is now effectively a no-op, since the numbers
2713 are always stored in normalized form. On a string, it creates a BigFloat
2718 Output values are BigFloat objects (normalized), except for bstr() and bsstr().
2720 The string output will always have leading and trailing zeros stripped and drop
2721 a plus sign. C<bstr()> will give you always the form with a decimal point,
2722 while C<bsstr()> (s for scientific) gives you the scientific notation.
2724 Input bstr() bsstr()
2726 ' -123 123 123' '-123123123' '-123123123E0'
2727 '00.0123' '0.0123' '123E-4'
2728 '123.45E-2' '1.2345' '12345E-4'
2729 '10E+3' '10000' '1E4'
2731 Some routines (C<is_odd()>, C<is_even()>, C<is_zero()>, C<is_one()>,
2732 C<is_nan()>) return true or false, while others (C<bcmp()>, C<bacmp()>)
2733 return either undef, <0, 0 or >0 and are suited for sort.
2735 Actual math is done by using the class defined with C<with => Class;> (which
2736 defaults to BigInts) to represent the mantissa and exponent.
2738 The sign C</^[+-]$/> is stored separately. The string 'NaN' is used to
2739 represent the result when input arguments are not numbers, as well as
2740 the result of dividing by zero.
2742 =head2 C<mantissa()>, C<exponent()> and C<parts()>
2744 C<mantissa()> and C<exponent()> return the said parts of the BigFloat
2745 as BigInts such that:
2747 $m = $x->mantissa();
2748 $e = $x->exponent();
2749 $y = $m * ( 10 ** $e );
2750 print "ok\n" if $x == $y;
2752 C<< ($m,$e) = $x->parts(); >> is just a shortcut giving you both of them.
2754 A zero is represented and returned as C<0E1>, B<not> C<0E0> (after Knuth).
2756 Currently the mantissa is reduced as much as possible, favouring higher
2757 exponents over lower ones (e.g. returning 1e7 instead of 10e6 or 10000000e0).
2758 This might change in the future, so do not depend on it.
2760 =head2 Accuracy vs. Precision
2762 See also: L<Rounding|Rounding>.
2764 Math::BigFloat supports both precision (rounding to a certain place before or
2765 after the dot) and accuracy (rounding to a certain number of digits). For a
2766 full documentation, examples and tips on these topics please see the large
2767 section about rounding in L<Math::BigInt>.
2769 Since things like C<sqrt(2)> or C<1 / 3> must presented with a limited
2770 accuracy lest a operation consumes all resources, each operation produces
2771 no more than the requested number of digits.
2773 If there is no gloabl precision or accuracy set, B<and> the operation in
2774 question was not called with a requested precision or accuracy, B<and> the
2775 input $x has no accuracy or precision set, then a fallback parameter will
2776 be used. For historical reasons, it is called C<div_scale> and can be accessed
2779 $d = Math::BigFloat->div_scale(); # query
2780 Math::BigFloat->div_scale($n); # set to $n digits
2782 The default value for C<div_scale> is 40.
2784 In case the result of one operation has more digits than specified,
2785 it is rounded. The rounding mode taken is either the default mode, or the one
2786 supplied to the operation after the I<scale>:
2788 $x = Math::BigFloat->new(2);
2789 Math::BigFloat->accuracy(5); # 5 digits max
2790 $y = $x->copy()->bdiv(3); # will give 0.66667
2791 $y = $x->copy()->bdiv(3,6); # will give 0.666667
2792 $y = $x->copy()->bdiv(3,6,undef,'odd'); # will give 0.666667
2793 Math::BigFloat->round_mode('zero');
2794 $y = $x->copy()->bdiv(3,6); # will also give 0.666667
2796 Note that C<< Math::BigFloat->accuracy() >> and C<< Math::BigFloat->precision() >>
2797 set the global variables, and thus B<any> newly created number will be subject
2798 to the global rounding B<immediately>. This means that in the examples above, the
2799 C<3> as argument to C<bdiv()> will also get an accuracy of B<5>.
2801 It is less confusing to either calculate the result fully, and afterwards
2802 round it explicitly, or use the additional parameters to the math
2806 $x = Math::BigFloat->new(2);
2807 $y = $x->copy()->bdiv(3);
2808 print $y->bround(5),"\n"; # will give 0.66667
2813 $x = Math::BigFloat->new(2);
2814 $y = $x->copy()->bdiv(3,5); # will give 0.66667
2821 =item ffround ( +$scale )
2823 Rounds to the $scale'th place left from the '.', counting from the dot.
2824 The first digit is numbered 1.
2826 =item ffround ( -$scale )
2828 Rounds to the $scale'th place right from the '.', counting from the dot.
2832 Rounds to an integer.
2834 =item fround ( +$scale )
2836 Preserves accuracy to $scale digits from the left (aka significant digits)
2837 and pads the rest with zeros. If the number is between 1 and -1, the
2838 significant digits count from the first non-zero after the '.'
2840 =item fround ( -$scale ) and fround ( 0 )
2842 These are effectively no-ops.
2846 All rounding functions take as a second parameter a rounding mode from one of
2847 the following: 'even', 'odd', '+inf', '-inf', 'zero' or 'trunc'.
2849 The default rounding mode is 'even'. By using
2850 C<< Math::BigFloat->round_mode($round_mode); >> you can get and set the default
2851 mode for subsequent rounding. The usage of C<$Math::BigFloat::$round_mode> is
2852 no longer supported.
2853 The second parameter to the round functions then overrides the default
2856 The C<as_number()> function returns a BigInt from a Math::BigFloat. It uses
2857 'trunc' as rounding mode to make it equivalent to:
2862 You can override this by passing the desired rounding mode as parameter to
2865 $x = Math::BigFloat->new(2.5);
2866 $y = $x->as_number('odd'); # $y = 3
2872 $x->accuracy(5); # local for $x
2873 CLASS->accuracy(5); # global for all members of CLASS
2874 # Note: This also applies to new()!
2876 $A = $x->accuracy(); # read out accuracy that affects $x
2877 $A = CLASS->accuracy(); # read out global accuracy
2879 Set or get the global or local accuracy, aka how many significant digits the
2880 results have. If you set a global accuracy, then this also applies to new()!
2882 Warning! The accuracy I<sticks>, e.g. once you created a number under the
2883 influence of C<< CLASS->accuracy($A) >>, all results from math operations with
2884 that number will also be rounded.
2886 In most cases, you should probably round the results explicitly using one of
2887 L<round()>, L<bround()> or L<bfround()> or by passing the desired accuracy
2888 to the math operation as additional parameter:
2890 my $x = Math::BigInt->new(30000);
2891 my $y = Math::BigInt->new(7);
2892 print scalar $x->copy()->bdiv($y, 2); # print 4300
2893 print scalar $x->copy()->bdiv($y)->bround(2); # print 4300
2897 $x->precision(-2); # local for $x, round at the second digit right of the dot
2898 $x->precision(2); # ditto, round at the second digit left of the dot
2900 CLASS->precision(5); # Global for all members of CLASS
2901 # This also applies to new()!
2902 CLASS->precision(-5); # ditto
2904 $P = CLASS->precision(); # read out global precision
2905 $P = $x->precision(); # read out precision that affects $x
2907 Note: You probably want to use L<accuracy()> instead. With L<accuracy> you
2908 set the number of digits each result should have, with L<precision> you
2909 set the place where to round!
2911 =head1 Autocreating constants
2913 After C<use Math::BigFloat ':constant'> all the floating point constants
2914 in the given scope are converted to C<Math::BigFloat>. This conversion
2915 happens at compile time.
2919 perl -MMath::BigFloat=:constant -e 'print 2E-100,"\n"'
2921 prints the value of C<2E-100>. Note that without conversion of
2922 constants the expression 2E-100 will be calculated as normal floating point
2925 Please note that ':constant' does not affect integer constants, nor binary
2926 nor hexadecimal constants. Use L<bignum> or L<Math::BigInt> to get this to
2931 Math with the numbers is done (by default) by a module called
2932 Math::BigInt::Calc. This is equivalent to saying:
2934 use Math::BigFloat lib => 'Calc';
2936 You can change this by using:
2938 use Math::BigFloat lib => 'BitVect';
2940 The following would first try to find Math::BigInt::Foo, then
2941 Math::BigInt::Bar, and when this also fails, revert to Math::BigInt::Calc:
2943 use Math::BigFloat lib => 'Foo,Math::BigInt::Bar';
2945 Calc.pm uses as internal format an array of elements of some decimal base
2946 (usually 1e7, but this might be different for some systems) with the least
2947 significant digit first, while BitVect.pm uses a bit vector of base 2, most
2948 significant bit first. Other modules might use even different means of
2949 representing the numbers. See the respective module documentation for further
2952 Please note that Math::BigFloat does B<not> use the denoted library itself,
2953 but it merely passes the lib argument to Math::BigInt. So, instead of the need
2956 use Math::BigInt lib => 'GMP';
2959 you can roll it all into one line:
2961 use Math::BigFloat lib => 'GMP';
2963 It is also possible to just require Math::BigFloat:
2965 require Math::BigFloat;
2967 This will load the necessary things (like BigInt) when they are needed, and
2970 Use the lib, Luke! And see L<Using Math::BigInt::Lite> for more details than
2971 you ever wanted to know about loading a different library.
2973 =head2 Using Math::BigInt::Lite
2975 It is possible to use L<Math::BigInt::Lite> with Math::BigFloat:
2978 use Math::BigFloat with => 'Math::BigInt::Lite';
2980 There is no need to "use Math::BigInt" or "use Math::BigInt::Lite", but you
2981 can combine these if you want. For instance, you may want to use
2982 Math::BigInt objects in your main script, too.
2986 use Math::BigFloat with => 'Math::BigInt::Lite';
2988 Of course, you can combine this with the C<lib> parameter.
2991 use Math::BigFloat with => 'Math::BigInt::Lite', lib => 'GMP,Pari';
2993 There is no need for a "use Math::BigInt;" statement, even if you want to
2994 use Math::BigInt's, since Math::BigFloat will needs Math::BigInt and thus
2995 always loads it. But if you add it, add it B<before>:
2999 use Math::BigFloat with => 'Math::BigInt::Lite', lib => 'GMP,Pari';
3001 Notice that the module with the last C<lib> will "win" and thus
3002 it's lib will be used if the lib is available:
3005 use Math::BigInt lib => 'Bar,Baz';
3006 use Math::BigFloat with => 'Math::BigInt::Lite', lib => 'Foo';
3008 That would try to load Foo, Bar, Baz and Calc (in that order). Or in other
3009 words, Math::BigFloat will try to retain previously loaded libs when you
3010 don't specify it onem but if you specify one, it will try to load them.
3012 Actually, the lib loading order would be "Bar,Baz,Calc", and then
3013 "Foo,Bar,Baz,Calc", but independent of which lib exists, the result is the
3014 same as trying the latter load alone, except for the fact that one of Bar or
3015 Baz might be loaded needlessly in an intermidiate step (and thus hang around
3016 and waste memory). If neither Bar nor Baz exist (or don't work/compile), they
3017 will still be tried to be loaded, but this is not as time/memory consuming as
3018 actually loading one of them. Still, this type of usage is not recommended due
3021 The old way (loading the lib only in BigInt) still works though:
3024 use Math::BigInt lib => 'Bar,Baz';
3027 You can even load Math::BigInt afterwards:
3031 use Math::BigInt lib => 'Bar,Baz';
3033 But this has the same problems like #5, it will first load Calc
3034 (Math::BigFloat needs Math::BigInt and thus loads it) and then later Bar or
3035 Baz, depending on which of them works and is usable/loadable. Since this
3036 loads Calc unnec., it is not recommended.
3038 Since it also possible to just require Math::BigFloat, this poses the question
3039 about what libary this will use:
3041 require Math::BigFloat;
3042 my $x = Math::BigFloat->new(123); $x += 123;
3044 It will use Calc. Please note that the call to import() is still done, but
3045 only when you use for the first time some Math::BigFloat math (it is triggered
3046 via any constructor, so the first time you create a Math::BigFloat, the load
3047 will happen in the background). This means:
3049 require Math::BigFloat;
3050 Math::BigFloat->import ( lib => 'Foo,Bar' );
3052 would be the same as:
3054 use Math::BigFloat lib => 'Foo, Bar';
3056 But don't try to be clever to insert some operations in between:
3058 require Math::BigFloat;
3059 my $x = Math::BigFloat->bone() + 4; # load BigInt and Calc
3060 Math::BigFloat->import( lib => 'Pari' ); # load Pari, too
3061 $x = Math::BigFloat->bone()+4; # now use Pari
3063 While this works, it loads Calc needlessly. But maybe you just wanted that?
3065 B<Examples #3 is highly recommended> for daily usage.
3069 Please see the file BUGS in the CPAN distribution Math::BigInt for known bugs.
3075 =item stringify, bstr()
3077 Both stringify and bstr() now drop the leading '+'. The old code would return
3078 '+1.23', the new returns '1.23'. See the documentation in L<Math::BigInt> for
3079 reasoning and details.
3083 The following will probably not print what you expect:
3085 print $c->bdiv(123.456),"\n";
3087 It prints both quotient and reminder since print works in list context. Also,
3088 bdiv() will modify $c, so be careful. You probably want to use
3090 print $c / 123.456,"\n";
3091 print scalar $c->bdiv(123.456),"\n"; # or if you want to modify $c
3097 The following will probably not print what you expect:
3099 my $c = Math::BigFloat->new('3.14159');
3100 print $c->brsft(3,10),"\n"; # prints 0.00314153.1415
3102 It prints both quotient and remainder, since print calls C<brsft()> in list
3103 context. Also, C<< $c->brsft() >> will modify $c, so be careful.
3104 You probably want to use
3106 print scalar $c->copy()->brsft(3,10),"\n";
3107 # or if you really want to modify $c
3108 print scalar $c->brsft(3,10),"\n";
3112 =item Modifying and =
3116 $x = Math::BigFloat->new(5);
3119 It will not do what you think, e.g. making a copy of $x. Instead it just makes
3120 a second reference to the B<same> object and stores it in $y. Thus anything
3121 that modifies $x will modify $y (except overloaded math operators), and vice
3122 versa. See L<Math::BigInt> for details and how to avoid that.
3126 C<bpow()> now modifies the first argument, unlike the old code which left
3127 it alone and only returned the result. This is to be consistent with
3128 C<badd()> etc. The first will modify $x, the second one won't:
3130 print bpow($x,$i),"\n"; # modify $x
3131 print $x->bpow($i),"\n"; # ditto
3132 print $x ** $i,"\n"; # leave $x alone
3134 =item precision() vs. accuracy()
3136 A common pitfall is to use L<precision()> when you want to round a result to
3137 a certain number of digits:
3141 Math::BigFloat->precision(4); # does not do what you think it does
3142 my $x = Math::BigFloat->new(12345); # rounds $x to "12000"!
3143 print "$x\n"; # print "12000"
3144 my $y = Math::BigFloat->new(3); # rounds $y to "0"!
3145 print "$y\n"; # print "0"
3146 $z = $x / $y; # 12000 / 0 => NaN!
3148 print $z->precision(),"\n"; # 4
3150 Replacing L<precision> with L<accuracy> is probably not what you want, either:
3154 Math::BigFloat->accuracy(4); # enables global rounding:
3155 my $x = Math::BigFloat->new(123456); # rounded immediately to "12350"
3156 print "$x\n"; # print "123500"
3157 my $y = Math::BigFloat->new(3); # rounded to "3
3158 print "$y\n"; # print "3"
3159 print $z = $x->copy()->bdiv($y),"\n"; # 41170
3160 print $z->accuracy(),"\n"; # 4
3162 What you want to use instead is:
3166 my $x = Math::BigFloat->new(123456); # no rounding
3167 print "$x\n"; # print "123456"
3168 my $y = Math::BigFloat->new(3); # no rounding
3169 print "$y\n"; # print "3"
3170 print $z = $x->copy()->bdiv($y,4),"\n"; # 41150
3171 print $z->accuracy(),"\n"; # undef
3173 In addition to computing what you expected, the last example also does B<not>
3174 "taint" the result with an accuracy or precision setting, which would
3175 influence any further operation.
3181 L<Math::BigInt>, L<Math::BigRat> and L<Math::Big> as well as
3182 L<Math::BigInt::BitVect>, L<Math::BigInt::Pari> and L<Math::BigInt::GMP>.
3184 The pragmas L<bignum>, L<bigint> and L<bigrat> might also be of interest
3185 because they solve the autoupgrading/downgrading issue, at least partly.
3187 The package at L<http://search.cpan.org/~tels/Math-BigInt> contains
3188 more documentation including a full version history, testcases, empty
3189 subclass files and benchmarks.
3193 This program is free software; you may redistribute it and/or modify it under
3194 the same terms as Perl itself.
3198 Mark Biggar, overloaded interface by Ilya Zakharevich.
3199 Completely rewritten by Tels L<http://bloodgate.com> in 2001 - 2006, and still