Thnaks a lot Jos. The idea of reading the array into a hash is quite appealing, and
simple too. But I have a small problem with this:
What if I want to sort on the second column of the array? Or if there are more
than two columns?
Say we have:
my @arr = qw(
1 2 3 4
4 5 6 7
5 6 7 8
1 2 3 4
3 1 8 3
);
And if I wanted to sort on the 3rd column.
Then I could write:
@sorted = sort { $a->[2] <=> $b->[2] } @arr;
# or 'cmp' if the sorting is to be lexical.
This should've worked. But why do I get a warning:
Use of uninitialized value at ./mk2_ratingchangedb.pl line 39, <F> chunk 8.
Here's the snippet of my code:
29 my(@dummy, @dummy2, @wed);
30 for ($i=1; $i<=$N; $i++) {
31 $dummy[$i][0] = &ParseDate($data{$key}[$i][0]);
32 if (! $dummy[$i][0]) {
33 warn "Could not parse $data{$key}[$i][0]\n";
34 }
35 $dummy[$i][1] = $data{$key}[$i][1];
36 }
37 @wed = @dummy;
38 &matrix_printer ("pre sort", @wed);
39 @dummy2 = sort { $a->[0] cmp $b->[0] } @wed;
40 printf " ----------------------------\n";
41 &matrix_printer ("post sort", @dummy2);
42 }
What could be wrong? ( The array @wed was REQUIRED. Otherwise my code
didn't work at all).
Thanks for the prompt response,
-tir
------------------------------------------------------------
On Sat, 16 Jun 2001, Jos I. Boumans wrote:
> ok, so if i get this right, @dummy has the following format:
>
> my @dummy = qw(
> 1996013100:00:00 MAAA
> 2000081100:00:00 MA-
> 1997063000:00:00 MAAA
> 1998122200:00:00 MAA
> 2000112400:00:00 MD
> );
>
> now we have that established, let's say that as soon as you think 'gee i
> need to SORT something on ONE VALUE in this data' you should probably be
> using a hash.
> so that's what we'll do
>
> what we'll do is read every line of @dummy, split it on white space and put
> that as key/value pairs in a hash, like so:
> my %hash = map { split /\s/ } @dummy;
>
> this is equivalent to the more verbose:
> my @data;
> for my $element (@dummy) {
> my ($key, $value) = split(/\s/, $element);
> push(@datay, $key, $value);
> }
> my %hash = @data;
>
> as you can see, the above is quite a bit more writing, slow and ugly then
> the map solution, but it's more 'logical' to read for most.
> also note that there's no objection to saying: %hash =
> @array_with_key_value_pairs
> the other way around is a bit more tricky, seeing hashes keep their
> key/value pairs seemingly unordened, wheras arrays keep indexes... so you
> can't say:
> %hash = @array_with_key_value_pairs;
> @copy_of_array_with_key_value_pairs = %hash;
> and assume that
> @copy_of_array_with_key_value_pairs eq @array_with_key_value_pairs;
>
> having said that, let's see how we can get our info out of the hash like we
> want to:
> for (sort keys %hash) {print "$_ has value $hash{$_}\n"}
>
> the above will print:
> 1996013100:00:00 is MAAA
> 1997063000:00:00 is MAAA
> 1998122200:00:00 is MAA
> 2000081100:00:00 is MA-
> 2000112400:00:00 is MD
>
> Hope this does what you want,
>
> hth,
>
> Jos Boumans
>
>
> <snipping stuff>
> > Folks,
> > # How do I sort an array by one of it's fields?
> >
> > # @dummy before sorting has this:
> >
> > 1996013100:00:00 MAAA
> > 2000081100:00:00 MA-
> > 1997063000:00:00 MAAA
> > 1998122200:00:00 MAA
> > 2000112400:00:00 MD
>
> > I would like to sort by it by the first column, which is a date read by
> > ParseDate fron Date::Manip.
> </snipping stuff>
>
>
Tirthankar C.Patnaik
Research Scholar
Indira Gandhi Institute of Development Research Ring:+91-22-840 0919/20/21
Goregaon East Off:- x593 Res: x675
Mumbai 400 065 Fax: 91-022-8402752
India. http://www.igidr.ac.in email:[EMAIL PROTECTED]
Never argue with an idiot. They drag you down to their level, then
beat you with experience. -dilbert
---------------------------------------------------------------------------
