Count similar characters in a row

[$new_guy]

Dear Monks,

I would like to count all the r's in the rows (row by row) then print to screen only the number of rows that have r's on all columns (i.e. 3 in this case). How do I do this. My script below doesn't seem to work well.

Thanks

script:

use strict;
use warnings;

# declare variables
my @column = ();
my $y;
my %hash;
my $count;
my $range = 3;

# open the file
open(INFILE,"<","my_column_file.txt") or die("Couldn't open the file\n
+");
      
     while (<INFILE>) {
        
        print "\n";                #separate the elements in each row 
+with a newline
               
        for ($y=0; $y < 3; $y++){   # from the first to the last field
+ (ie field/column 3)
            
         my ($field) = (split)[$y]; # extract fields/ columns from the
+ data
         push @column, $field;      # store all the field values in an
+ a array
         
         #count the consecutive r's --> all those common to a row
         #(i.e.if a sinlge specified column misses an r at that row do
+n't cont it)!

          while ($field =~ /(r+)/g) {
            my $new_count = $count++;
            print "$new_count\n";
            }
                       
           print pop@column;       # print from the array, each time r
+emoving the printed element (avoiding repetition)
           print "\n";             # print each element in a new line
            
    }  
}

my file:

0  a b h
1  - r z
3  u - u
4  r x r
5  r t r
6  r r r
7  r r r
8  r r r

[toolic]

use strict;
use warnings;

while (<DATA>) {
   my @cols = split;
   shift @cols;
   my $all_r = 1;
   for my $c (@cols) {
      if ($c ne 'r') {
         $all_r = 0;
         last;
      }
   }
   print if $all_r;
}

__DATA__
0  a b h
1  - r z
3  u - u
4  r x r
5  r t r
6  r r r
7  r r r
8  r r r

prints out:

6  r r r
7  r r r
8  r r r

update: even better with List::MoreUtils:

use strict;
use warnings;
use List::MoreUtils qw(all);

while (<DATA>) {
   my @cols = split;
   print if all { $_ eq 'r' } @cols[1 .. $#cols];
}

__DATA__
0  a b h
1  - r z
3  u - u
4  r x r
5  r t r
6  r r r
7  r r r
8  r r r

[johngg]

An alternative using tr (see Transliteration in Quote and Quote like Operators) to count the number of 'r's.

knoppix@Microknoppix:~$ perl -E '
> open my $fh, q{<}, \ <<EOD or die $!;
> 0  a b h
> 1  - r z
> 3  u - u
> 4  r x r
> 5  r t r
> 6  r r r
> 7  r r r
> 8  r r r
> EOD
> 
> while ( <$fh> )   
> {
>     my( $seq, @cols ) = split;
>     print if tr{r}{} == scalar @cols;
> }'
6  r r r
7  r r r
8  r r r
knoppix@Microknoppix:~$

[$new_guy]

Thanks Monks,

They all work well. However, except the perl one liners suggested no other scripts does what I wanted which in the end was to count the number ie 3 and print it on the screen!

Also, how do I adapt the scripts to do this for a list (a second file - random.txt) separated by space as shown in the example below:

2

3
2

3
2
1

2

1
3

2
1
3

3

3
1

3
1
2

Notice there is a space in the first line. The numbers are clustered in the manner in which the r's common to all columns should be counted. For example print only the r's in column 2, then print all the r's common in all rows of column 3 and 2, then print the r's common to all rows in column 3 , 2 and 1.

so the printed output for my file my_column_file.txt should below:

0  a b h
1  - r z
3  u - u
4  r x r
5  r t r
6  r r r
7  r r r
8  r r r 

<p>should be:(the out put with explainations) </p>



4 3 3  (ie for 2 32 321)
4 5 3  (ie for 2 13 213)
5 5 3  (ie for 3 31 312)

the out put desired 

4 3 3  
4 5 3  
5 5 3

[johngg]

Addressing your first issue is simple a matter of incrementing a count rather than printing the line.

$ perl -Mstrict -wE '
> open my $fh, q{<}, \ <<EOD or die $!;
> 0  a b h
> 1  - r z
> 3  u - u
> 4  r x r
> 5  r t r
> 6  r r r
> 7  r r r
> 8  r r r
> EOD
> 
> my $count;
> while ( <$fh> )
> {
>     my( $seq, @cols ) = split;
>     $count ++ if tr{r}{} == scalar @cols;
> }
> say $count;'
3
$

The supplementary question requires a little more jiggery-pokery.

$ perl -Mstrict -wE '
> open my $rand, q{<}, \ <<EOD or die $!;
> 2 32 321
> 2 13 213
> 3 31 312
> EOD
> 
> my @tests =
>    map [ split ],
>    <$rand>;
> 
> open my $fh, q{<}, \ <<EOD or die $!;
> 0  a b h
> 1  - r z
> 3  u - u
> 4  r x r
> 5  r t r
> 6  r r r
> 7  r r r
> 8  r r r
> EOD
> 
> my @results;
> while ( <$fh> )
> {
>     my @cols = split;
>     foreach my $idx ( 0 .. $#tests )
>     {
>         foreach my $subidx ( 0 .. $#{ $tests[ $idx ] } )
>         {
>             my @posns = split m{}, $tests[ $idx ]->[ $subidx ];
>             $results[ $idx ]->[ $subidx ] ++
>                if scalar @posns == grep { q{r} eq $cols[ $_ ] } @pos
+ns;
>         }
>     }
> }
> 
> say qq{@$_} for @results;'
4 3 3
4 5 3
5 5 3

I hope this is useful.

[suhailck]

Try this,



perl -wle 'while(<>) { $count++ if "rrr" eq join "",(split)[1 .. 3] };
print $count' my_column_file.txt

3

[jwkrahn]

perl -lane '$count++ if "rrr" eq join "",@F[1 .. 3] }{ print $count' m
+y_column_file.txt

[jwkrahn]

# open the file
open INFILE, '<', 'my_column_file.txt' or die "Couldn't open 'my_colum
+n_file.txt' $!";

while ( <INFILE> ) {
    my ( $row_num, @fields ) = split;
    print if @fields == grep /r/, @fields;
    }