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

Okay, so here's the deal
I need a data structure that will store 2-7 data points (these points will be numbers from 0-50 with none repeated) in such a way as to preserve the order of the data points, and to make look up on a specific point easy.

I'm going to address the second part first, to make this easy for me, I'm tempted to put these items into a hash and use exists. eg.
sub find_point(){
    my $ref = shift;
    my %hash = %$ref;
    my $key = shift;
    if(exists($hash{$key})){
        ...
    }
}

[download]

While this would be an ideal solution for finding a key quickly, trying to order the data based on the values associated with those keys seems to be quite a bit of extra effort. Eg.
sub points_list(){
    my $ref = shift;
    my %hash = %$ref;
    my @order;
    foreach $item (keys %hash){
        @order[int($hash{$item})]=$item;
    }
    return \@order;
}

[download]

Now this wouldn't be so bad, but it may be done on a number of hashes like this. Does anyone see a better way?
==
Kwyjibo. A big, dumb, balding North American ape. With no chin.

Replies are listed 'Best First'.
Re: I Ordered My Keys With Existing Values!
by hardburn (Abbot) on Jul 28, 2003 at 15:03 UTC 
    my @order = sort {
   $hash{$a} <=> $hash{$b}
} keys %hash;

    [download]

    Or you can try Tie-InsertOrderHash-0.01.

    ----
    I wanted to explore how Perl's closures can be manipulated, and ended up creating an object system by accident.
    -- Schemer

    Note: All code is untested, unless otherwise stated

[reply]
[d/l]

      I thought that was just Tie::IxHash which is more common.

[reply]

        *shrug* I did a search on CPAN for "order hash tie", and the InsertOrder module came up first. I've never found a use for keeping the insertion order on hashes, so I don't really have an opinion on either one.

        ----
        I wanted to explore how Perl's closures can be manipulated, and ended up creating an object system by accident.
        -- Schemer

        Note: All code is untested, unless otherwise stated

[reply]
      This is an interesting piece of code, and while it is doing exactly what I'm doing, it doesn't address the underlying problem that I was trying to represent.

      Having to do this every time I want to display this hash will require quite a bit of extra calculations. I could store this array but then I have data duplication (memory/speed tradeoff) that I was hoping to avoid.
      ==
      Kwyjibo. A big, dumb, balding North American ape. With no chin.

[reply]
        With 2 to 7 data points the extra calculations are something less than "quite a bit", unless you're displaying the values many, many times. I suspect the threshold would be quite high before you noticed a slowdown.
[reply]
Re: I Ordered My Keys With Existing Values!
by traveler (Parson) on Jul 28, 2003 at 15:30 UTC
    Two more options: 1) use both an array and a hash. While I generally eschew such parallel data structures, this might be a good use; 2) add a hash key specifying the order then use that in a loop to print out the values.

    I dunno which is better for you. It likely depens on how often you access the "sorted" values and how many hashes you end up having.

    HTH, --traveler

[reply]
Re: I Ordered My Keys With Existing Values!
by TomDLux (Vicar) on Jul 28, 2003 at 17:53 UTC

    In find_point() you pass in a reference to a hash, and then copy the hash from %$ref to %hash. If you're concerned about speed, dedreference the reference: 

    if( exists $ref->{$key} ) {

    [download]

    Since the hashes are tiny, 2 to 7 members, copy it from the beginning, rather than passing in a referernce.

    --
    TTTATCGGTCGTTATATAGATGTTTGCA

[reply]
[d/l]

Back to Seekers of Perl Wisdom