http://www.perlmonks.org?node_id=1020253

cpaplham has asked for the wisdom of the Perl Monks concerning the following question:

I'm trying to add one hundred 50-digit numbers together. my code is basically:
my $big_number = 37107287533902102798797998220837590246510135740250 + 46376937677490009712648124896970078050417018260538 + ...;
I have tried  use bignum; but this didn't work. Whenever I print the value, I get an empty string. I only need about 15 digits of precision.

Replies are listed 'Best First'.
Re: Adding really big numbers
by frozenwithjoy (Priest) on Feb 23, 2013 at 00:15 UTC
    Which version of Perl are you using? It works just fine for me on 5.16.1. Using bigint, I get values for every integer. Using bignum, I get values for every int and decimal (except for trailing zeroes). Without an extra module, anything over 20 digits defaults to scientific notation with 15 digits of precision.

    Update: Here is some info on largest positive/negative integers that can be nadled by different setups w/o using modules: Re: maximum value of a scalar

      I'm using 5.14. I agree, it's strange.
        Odd. Works fine for me on 5.14.2. What is your operating system?
Re: Adding really big numbers
by salva (Canon) on Feb 23, 2013 at 09:33 UTC
    If you need to do that application wide, you can compile a custom perl that uses long doubles to represent floating point numbers.

    x86 long doubles have 80 bits and a resolution of 18 digits. Other architectures as SPARC use 128 bits for long doubles.

    Otherwise, you can use Math::LongDouble or Math::MPFR

Re: Adding really big numbers
by Khen1950fx (Canon) on Feb 23, 2013 at 01:58 UTC
    What lib are you using? How are you declaring accuracy and precision? Maybe this'll help:
    #!/usr/bin/perl -l BEGIN { $| = 1; $^W = 1; } use strict; use warnings; use bignum lib => 'FastCalc'; use Memoize; memoize('really_big'); really_big(); sub really_big { bignum->accuracy(15); bignum->precision(15); my $big_number = 3710728753390210279879799822083759024651013574025 +0 + 46376937677490009712648124896970078050417018260538; print $big_number; }
Re: Adding really big numbers
by ambrus (Abbot) on Mar 13, 2013 at 21:34 UTC

    If you only need about 15 digits of precision, you don't need to do anything special. Perl's built in arithmetic will read the numbers as double-precision floats and add them that way, which gives about 15 digits of precision.