random numbers

[wahrana]

I have a problem that I can't seem to solve. So, maybe there is someone out there who can help.

Here is my problem. I have a file with a certain number of entries in it. Each entry is a one line in the file. I have a certain number of users. How can I write a Perl code that will divide the number of entries randomly between the users. Every time the code is executed, a new set is generated. For example, I have a file with 20 entries and 5 users. Dividing equally, each user gets 4 entries. Divided randomly,
users 1 gets 8, user 2->3, user 3->3, user 4->2 and user 5->4. Next execution something different, but the total of entries is always equal to 20. And all the users get a share.

Thanks.

[KevinADC]

What have you tried so far?

[miller]

There are lots of different ways to do this. What have you tried so far?

Obviously this will take some type of algorithm. But any solution will probably simply need one of the following resources:

perldoc rand
cpan List::Util - The shuffle function

- Miller

[KevinADC]

Usually I don't post code unless you have shown some effort, but the question was interesting enough that I thought about it and came up with a very simple method:

my @users = qw(john joe jane jerry janet);
my @data = (foo) x 20;
my %users = map {$_ => []} @users;
for (@data) {
   push @{$users{$users[int rand @users]}},$_;
}
foreach my $user (@users) {
   print "$user = @{$users{$user}}\n";
} 

This works by picking a random user to get some data each time through the "for" loop. There is no need to check how many users there are or how much data there is. This might not actually be what you want because sometimes a user can have zero data and there is a very very small possibility that one user could get all the data. But it would be easy to make sure each user had at least one bit of data or check for a maximum amount of data per user.

[miller]

I agree about this type of problem being interesting. Here's a slight modification to the code that you provided that balances out the distribution across all users.

my @users = qw(john joe jane jerry janet);
my @data = (1..20);
my %users = map {$_ => []} @users;
for (my $i = 0; @data; $i = ($i + 1) % @users) {
    push @{$users{$users[$i]}}, splice @data, int(rand @data), 1;
}
foreach my $user (@users) {
    print "$user = @{$users{$user}}\n";
}
 

Typical output:

>perl scratch.pl
john = 15 3 14 8
joe = 5 4 1 12
jane = 19 11 18 20
jerry = 9 10 17 6
janet = 13 2 7 16
 

- Miller

[KevinADC]

Cool code. :)


........................................