That introduces an error of rounding up when it should be down.
The only way I can com up with for my test case is to replace that line with:
$number = abs($number)*$dec +.5;
$number =~ s/\..*$//;
which instead of using int to truncate which did not work in my test case, treats it as a sting and truncates the decimal that way.

So, the sub now is:
sub myRound {
my $number = shift || 0;
my $places = shift || 0;
my $dec = 10**$places;
my $sign = ($number >= 0) ? 1 : -1;

#$number = sprintf(”%.0f”,(abs($number)*$dec) +.5);
$number = abs($number)*$dec +.5;
$number =~ s/\..*$//;

# return the number with sign and correct decimal places
return $number * $sign / $dec;
}

my $calc = round(1900*18.12895,2)
This returned 34445
The correct value should be 34445.01

This might be due to the fact that int sometimes doesn’t behave as expected (see http://perldoc.perl.org/functions/int.html)
Also, I got strange results from ($number 0) which is supposed to handle negative numbers.

Here is the modification I came up with:

sub myRound {
my $number = shift || 0;
my $places = shift || 0;
my $dec = 10**$places;
my $sign = ($number >= 0) ? 1 : -1;

$number = sprintf(”%.0f”,(abs($number)*$dec) +.5);

# return the number with sign and correct decimal places
return $number * $sign / $dec;
}

This gave me the correct result, and it also handles negative numbers:
my $calc – myRound(1900*-18.12895,2)
returns -34445.01

Thanks again!

print round(3.14159265,3); # ==> 3.142

sub round {
my $nombre = shift || 0;
my $dec = 10 ** (shift || 0);
#return sprintf(“%.${dec}f”, $nombre);
return int( $dec * $nombre + .5 * ($nombre 0)) / $dec;
}