Beefy Boxes and Bandwidth Generously Provided by pair Networks
P is for Practical
 
PerlMonks  

dynamic map "quadrant" indexing

by Anonymous Monk
on Aug 14, 2005 at 09:35 UTC ( #483673=perlquestion: print w/ replies, xml ) Need Help??
Anonymous Monk has asked for the wisdom of the Perl Monks concerning the following question:

Hello Monks,

this is so embarassing. I feel this is easy, but somehow I'm too stupid to achieve it. Needed is a subroutine, that will print letter/number indices of map quadrants. Depending on the level (zoom):

&idx(1); # will spill out: A1 B1 C1 D1 &idx(2); # will spill out A1A1 A1B1 A1C1 A1D1 # B1A1 B1B1 B1C1 B1D1 # C1A1 C1B1 C1C1 C1D1 # D1A1 D1B1 D1C1 D1D1 &idx(3); # should do A1A1A1 A1A1B1 A1A1C1 A1A1D1 # A1B1A1 A1B1B1 A1B1C1 A1B1D1 # etc.

Any means of doing this iteratively failed miserably, probably a recursive solution would be best, but somehow I'm failing to get a grip on the numbers:

$lvl can be anything from 1 to n, which results in 4**$lvl return values, each value consisting of $lvl <char><digit> numbers. As you probably figured out, chars are always [A-D] and digits [1-4].

This probably wouldn't be used for $lvl > 6, but you never know so I wanted to go for a generic solution.

Do you have any idea HowTo?

Thanks a lot,
Marcel

Comment on dynamic map "quadrant" indexing
Select or Download Code
Re: dynamic map "quadrant" indexing
by Anonymous Monk on Aug 14, 2005 at 10:20 UTC
    Ok - I did it. Wasn't that hard after all, and now I'll make the code more elegant.
    use strict; my $lvl ="@ARGV"; my $out = &idx($lvl); print "@$out\n"; sub idx { my $lvl = shift; my @out = qw(a1 b1 c1 d1); if($lvl == 1) { return \@out; } else { my @tmp = (); for my $o (@out) { push @tmp, "$o$_" for(@{&idx($lvl-1)}); } @out = @tmp; } return \@out; }
    Marcel
      Uhm... well forget it, the code is crap. This one does the right thing:
      sub idx { my $lvl = shift; # ARG1: get level my @out = qw(a1 b1 c1 d1 a2 b2 c2 d2 a3 b3 c3 d3 a4 b4 c4 d4); # init OUT array return \@out if($lvl == 1); # bail out if lvl = 1 my @tmp; for my $o (@out) { push @tmp, map { "$o$_" } @{&idx($lvl-1)}; } return \@tmp; }
      Probably I should not do my coding sunday morning. Marcel
Re: dynamic map "quadrant" indexing
by tlm (Prior) on Aug 14, 2005 at 14:46 UTC

    Update: I think the glob-based solution in my other reply is the way to go; I wish I'd thought of it first. Ignore this one.


    Here's a recursive solution:

    the lowliest monk

Re: dynamic map "quadrant" indexing
by tlm (Prior) on Aug 14, 2005 at 15:35 UTC

    Here's an iterative solution, similar in spirit to the algorithm discussed in HOP, flip, and swap:

    But it's much easier to do this with globs:

    my $glob_string = ( '{' . join( ',', 'A'..'D' ) . '}1' ) x $level; my @indices; while ( < $glob_string > ) { push @indices, $_ }

    the lowliest monk

      Or with Algorithm::Loops:
      use Algorithm::Loops qw( NestedLoops ); my $level = 3; my @indices = NestedLoops( [ (['A'..'D']) x $level ], sub { join '', map { "${_}1" } @_ } ); foreach my $val (@indices) { print($val\n") ]
      or as an interator:
      use Algorithm::Loops qw( NestedLoops ); my $level = 3; my $iter = NestedLoops( [ (['A'..'D']) x $level ] ); while (my @values = $iter->()) { my $val = join '', map { "${_}1" } @values; print("$val\n"); }
      Very nice ++

      I did not know about this glob behavior. I had to change it a little to make it work exaktly like the op wanted though.
      my $level = 2; my $l = 'A'; for (2 .. $level){ $l++; } my $glob_string = ( '{' . join( ',', 'A'.. $l ) . '}1' ) x $level; my @indices = glob $glob_string; print "@indices\n";

Log In?
Username:
Password:

What's my password?
Create A New User
Node Status?
node history
Node Type: perlquestion [id://483673]
Approved by neniro
Front-paged by planetscape
help
Chatterbox?
and the web crawler heard nothing...

How do I use this? | Other CB clients
Other Users?
Others examining the Monastery: (21)
As of 2014-07-29 14:02 GMT
Sections?
Information?
Find Nodes?
Leftovers?
    Voting Booth?

    My favorite superfluous repetitious redundant duplicative phrase is:









    Results (217 votes), past polls