What does your existing code look like? It could be that your implementation, using a hash (which seems entirely reasonable) is almost on the mark. Wouldn't it be of more value to you for us to help figure out what you're doing wrong, rather than having us provide you a cut-and-paste solution?

Here's how I would do it: Create a hash. Then iterate over each line of the file. chomp each line (to get rid of the \n). split on the ':' character. Retain the first element (state) and the third element (temperature). push each temperature into an anonymous array held as the value of the hash element indexed by the state: push @{$states{$state}}, $temp;. Then open your output file. Iterate over each state in the hash. Calculate the average by summing all elements for a given state, and then dividing by the number of elements. Save to your output file the state name and the average temperature. close your output file. Be sure to check return values on filesystem calls such as open and close.

"I'm using hash" is not illustrative enough of the code you've tried so far. Is this your homework? Do you want us to do it for you?

Hints: first, learn about split. Then, suppose you somehow you've managed to obtain a %temp hash, which happens to be a "HoA"; you probably will do something like:

for (keys %temp) {
    my $tot;
    $tot += $_ for @{ $temp{$_} };
    print "Average for state '$_': ", $tot/@{ $temp{$_} }, "\n";
}
[download]

but possibly with additional checks and taylored to your needs. Suggested readings: perlref and the other docs mentioned there under the heading "SEE ALSO".

perl -F: -wane '$t{$F[0]}+=$F[2]; $c{$F[0]}++; 
                END{printf "%s: %f\n", $_, $t{$_}/$c{$_} 
                    for keys %c}' file
[download]

The solution looks accurate, although, especially being terse, some added explanation for this newbie OP, or perhaps where to look in the docs, seems warranted.

-M

Free your mind

Terse, cryptic, partially golfed answers are typically acceptable if the question smells like homework. Consider such to be defense mechanisms of the monestary; good answers to homework questions will attract more homework questions. I, and many others, believe many homework questions would decrease the value of this community.

-Scott

some added explanation for this newbie OP, or perhaps where to look in the docs, seems warranted.

If you think that it is warranted, what's keeping you from posting it?
Perl --((8:>*

my %HoH = (); # 'my' declares at file scope in this case
LoadFile(\%HoH); # passing by reference is more portable
                 # if you later make a module
Stats(\%HoH);

sub LoadFile{
    my $href = shift; # get the actual parameter
    open my $fileHandle, "<filepath" or die $!;
    while( <$fileHandle> {
        my ($st,$to,$te) = split( /\:/ );
        $href -> { $st }{ $to } = $te; # load file data into the HoH
    }
    close $fileHandle;
}

sub Stats {
    my $href = shift;
    for my $st ( sort keys %$href ) { # in state order
        my $toref = $href -> { $st }; # ref. to town subhash
        my ( $count, $sum ) = (0,0);
        for my $to ( keys %$toref ) { # sum through towns
                                      # (any order will do)
            $count++; $sum += ( $toref -> { $to } );
                # accumulate state aggregates
        }
        print "$st: " . $sum/$count . "\n";
            # average for state = sum/count
    }
}
[download]