The way you set up the arrays is very uncommon and may not be what you want. The data structure you choose depends in part in what you want to do with the results. Just print them out to a report, perform calculations on them, etc. By looking at the fields you're capturing, these seem to be business statistics. It would be helpful to know what it is you want to do with the results. If each of the variables you set up equaled the number of returned fields, there are about 21 of them.

I am probably making this more difficult then nescessary. I have been leveraging code done by a previous programmer, as I am very new to PERL.

I welcome better ideas on how to get the desired output, which is an final excel CSV.

The avgerage cost is to be used to preform calulations in the second "array" .

I suspect a foreach loop next to pull in average cost and caluculate new fields. Caluclate a couple of other fields based on available data and then output to CSV.

Well, here's a little to get started ;-)

I am guessing, (from the code you have inherited), that you want the rows returned from the database query reported as columns - not rows. I wrote a transpose function for that. If that's not the case, then don't transpose @data.

This code could be a little sticky for a beginner, so ask questions on what you don't understand.

Note that the @col array contains the 'headers' of the data you want (in rows or columns, I don't know what is required).

These columns should be the names of the columns returned by the query (THAT you intend to be in the report) plus any additional fields you want to include, (mkt_1, profit, percent_profit, etc..).

my %ave12_cost;

while (my @row = $sth_2->fetchrow_array) {
    
    # @row has 2 items - item AND average cost for the item
    
    # remove trailing spaces    
    s/\s+$// for @row;

    # save item and average cost in a hash whose key is 'item'
    # 123.4 becomes 123.40 AND 456 becomes 456.00
    $ave12_cost{ $row[0] } = sprintf "%.2f", $row[1];
}

$sth_2->finish();

my @col = qw(
agc
market
mkt_1
analysis_code
item
customer
key_val2
key_val3
terms_code
audit_seq
aud_date
aud_time
opr_id
action
unit_price
profit
percent_profit
cur_per_month
);

my @data = [ @col ];

while (my $href = $sth_3->fetchrow_hashref) { 

    # remove trailing spaces    
    s/\s+$// for values %$href;
    
    # do calculations    
    $href->{mkt_1} = substr($href->{market}, 0, 1) == 1 ? "A&D" : "MCP
+";
    $href->{profit} = $href->{unit_price} - $ave12_cost{$href->{item}}
+;
    $href->{percent_profit} = sprintf "%.1f", ($href->{profit} / $href
+->{unit_price}) * 100;
    . . . .
    
    # add to an array for output to Excel
    # (Data added in order of '@col' array)
    push @data, [ @$href{ @col } ];
}

$sth_3->finish();

my @transposed = transpose( @data );

my $file = 'sales.csv';
open my $out, ">", $file or die "Unable to create $file. $!";

# print to CSV file
for my $row (@transposed) {
    print $out join(",", @$row), "\n";
}

close $out or die "Unable to close $file. $!";


sub transpose {
    my (@data, @trans) = @_;
    for my $r (0 .. $#data) {
        for my $c (0 .. $#{ $data[$r] }) {
            $trans[$c][$r] = $data[$r][$c];
        }
    }
    return @trans;
}
[download]

Just out of curiosity, why are you rewriting the code? Does it still work, (even if it's not the best/most pretty)?

If there are embedded commas in any of the fields, the print to the file will fail. The statement print $out join(",", @$row), "\n"; would have to be replaced by a method from a module like Text::CSV_XS.