«

Hi,

I'm only familiar with two programming environments. That's Perl and
SPSS (Statistical Package for the Social Science, www.spss.com). This
is actually an SPSS question, but I got no answer from
comp.soft-sys.stat.spss so I turn to this ng since it's the other
environment that I understand (at least roughly). And I can pretty
much translate my SPSS code to Perl.

I have tree lists:

@rows=('q2','q8','q10','q11','q19','q20','q22','q2 9');
@m_vars=('q10','q19','q20');
@m_labs=('Filter A','Filter B','Filter C');

and I want this print:

q2
q8
Filter A
q11
Filter B
Filter C
q22
q29

I want to loop through @rows, and whenever I encounter a match between
an item in @rows and an item in @m_vars, I want to print the
corresponding item in @m_labs.

Here's the catch: In SPSS I only have access to a foreach-like loop. I
can use a if-structure with "next" ("last", "redo" and "goto" are not
available). The "elsif" is not available in SPSS if-structure, only
"else".

I have a work-around for an auto increment like thingy (++).

With all the tools available in Perl, this is probably a no-brainer,
but the constraints replicated from SPSS suggests a robust solution...
Anyone?

This the translated SPSS code I have so far:

$match=0;
@rows=('q1','q2','q3','q4','q5','q6','q7','q8');
@m_vars=('q3','q5','q6');
@m_labs=('Filter A','Filter B','Filter C');

foreach $i (@rows) {
$i_count=1;
$j_count=1;
foreach $j (@m_vars) {
$k_count=1;
foreach $k (@m_labs){

# Set a flag to use in the
# last condition below...
if($i eq $j && $k_count == 1){
$match++;
}

# Print the main list
if ($i ne $j && $i_count == 1) {
print "$i\n";
}

# Print the "Filters"...
if ($i eq $j && $match == $k_count) {
print "$k\n";
}
$i_count++;
$k_count++;
}
$j_count++;
}
}

On Wed, 21 Jul 2004 23:32:34 -0700, petersson wrote:
> I have tree lists:
>
> @rows=('q2','q8','q10','q11','q19','q20','q22','q2 9');
> @m_vars=('q10','q19','q20');
> @m_labs=('Filter A','Filter B','Filter C');
>
> and I want this print:
>
> q2
> q8
> Filter A
> q11
> Filter B
> Filter C
> q22
> q29
>
> I want to loop through @rows, and whenever I encounter a match between
> an item in @rows and an item in @m_vars, I want to print the
> corresponding item in @m_labs.

Here's a suggestion for you; the idea is to convert @m_vars to hash where
the element value is the key and the element's position is the value. I
haven't tested it, though;

#!/usr/bin/perl
#
use strict;
use warnings;

my @rows = ('q2','q8','q10','q11','q19','q20','q22','q29');
my @m_vars = ('q10','q19','q20');
my @m_labs = ('Filter A','Filter B','Filter C');

## Convert @m_vars to a hash
my %m_hash = ();
for ( my $i = 0; $i <= $#m_vars; $i++ ) {
$m_hash{$m_labs[$i]} = $i;
}

## Output
foreach ( @rows ) {
if ( exists $m_hash{$_} ) {
print $m_labs[$m_hash{$_}] . "\n";
}
else {
print $_ . "\n";
}
}


--
Tore Aursand <>
"There are three kinds of lies: lies, damn lies, and statistics."
(Benjamin Disraeli)

(petersson) wrote in message news:<. com>...
> Hi,
>
> I'm only familiar with two programming environments. That's Perl and
> SPSS (Statistical Package for the Social Science, www.spss.com). This
> is actually an SPSS question, but I got no answer from
> comp.soft-sys.stat.spss so I turn to this ng since it's the other
> environment that I understand (at least roughly). And I can pretty
> much translate my SPSS code to Perl.
>
> I have tree lists:
>
> @rows=('q2','q8','q10','q11','q19','q20','q22','q2 9');
> @m_vars=('q10','q19','q20');
> @m_labs=('Filter A','Filter B','Filter C');
>
> and I want this print:
>
> q2
> q8
> Filter A
> q11
> Filter B
> Filter C
> q22
> q29
>
> I want to loop through @rows, and whenever I encounter a match between
> an item in @rows and an item in @m_vars, I want to print the
> corresponding item in @m_labs.
>
> Here's the catch: In SPSS I only have access to a foreach-like loop. I
> can use a if-structure with "next" ("last", "redo" and "goto" are not
> available). The "elsif" is not available in SPSS if-structure, only
> "else".
>
> I have a work-around for an auto increment like thingy (++).
>
> With all the tools available in Perl, this is probably a no-brainer,
> but the constraints replicated from SPSS suggests a robust solution...
> Anyone?
>
> This the translated SPSS code I have so far:
>
> $match=0;
> @rows=('q1','q2','q3','q4','q5','q6','q7','q8');
> @m_vars=('q3','q5','q6');
> @m_labs=('Filter A','Filter B','Filter C');
>
> foreach $i (@rows) {
> $i_count=1;
> $j_count=1;
> foreach $j (@m_vars) {
> $k_count=1;
> foreach $k (@m_labs){
>
> # Set a flag to use in the
> # last condition below...
> if($i eq $j && $k_count == 1){
> $match++;
> }
>
> # Print the main list
> if ($i ne $j && $i_count == 1) {
> print "$i\n";
> }
>
> # Print the "Filters"...
> if ($i eq $j && $match == $k_count) {
> print "$k\n";
> }
> $i_count++;
> $k_count++;
> }
> $j_count++;
> }
> }


This might be simpler (although not extremely perlish):

#!/usr/bin/perl
use strict;
#
my @rows=('q2','q8','q10','q11','q19','q20','q22','q2 9');
my @m_vars=('q10','q19','q20');
my @m_labs=('Filter A','Filter B','Filter C');
my $out;
my ($i, $j);

for ($i=0;$i<@rows;$i++) {
$out = $rows[$i];
for ($j=0;$j<@m_vars;$j++) {
$out = $m_labs[$j] if ($rows[$i] eq $m_vars[$j]);
}
print $out, "\n";
}