How can I set the level of precision for Perl's bignum?

Question

I'm trying to use the bignum module in Perl and want to set the precision. I know this can be done via a one liner as detailed on the module's CPAN page:

$ perl -Mbignum=p,-50 -le 'print sqrt(20)'

...which will print out the square root of 20 to 50 digits of precision, but what I'm wondering is if there's anyway to set the precision within a script, i.e. something like:

#!/usr/bin/perl
use bignum;

setPrecision(-50);
print sqrt(20);

I've searched around here, Google, and PerlMonks without any luck so far. Thanks in advance.

Specify it as an argument to the package in the `use bignum` line. — Anon., Jan 25 '10 at 23:01
See `perldoc perlrun`; `perl -MPackage=foo,bar` is equivalent to starting the program with `use Package qw(foo bar)`. — ephemient, Jan 25 '10 at 23:13
Related: http://stackoverflow.com/questions/1838808/how-do-i-set-the-floating-point-precision-in-perl — martin clayton, Jan 25 '10 at 23:55

score 12 · Accepted Answer · edited Jan 26 '10 at 00:02

12

Per Anon.'s suggestion:

#!/usr/bin/perl

use strict;
use warnings;

use bignum ( p => -50 );

print sqrt(20);

You might like to look at the docs for Math::BigFloat and Math::BigInt which bignum makes use of.

edited Jan 26 '10 at 00:02

Brad Gilbert

33,846
11
78
129

answered Jan 25 '10 at 23:06

martin clayton

76,436
32
213
198

1

That does not work for me, Perl is wrong here: `perl -e 'use bignum(p=>-100);printf("%1.50f\n",sqrt(23));'` outputs `4.79583152331271911350540904095396399497985839843750` This is right: `perl -Mbignum=p,-50 -le 'print sqrt(23)'` `4.79583152331271954159743806416269391999670704190413` compared with the correct calculation from bc: `echo "scale=50;sqrt(23)"|bc` `4.79583152331271954159743806416269391999670704190412` – Axel Heider Oct 24 '13 at 10:03

How can I set the level of precision for Perl's bignum?

1 Answers1