Find out range for every field in a column - indexing

Question

I have a column that looks like this:
A 1
B 3
C 5
D 4
E 7
F 1
G 1
H 3
For every filed in column#2, I want to calculate the range (max-min) of 3 field radius up and down.
A range(1 3 5 4)
B range(1 3 5 4 7)
C range(1 3 5 4 7 1)
D range(1 3 5 4 7 1 1)
E range( 3 5 4 7 1 1 3)
F range( 5 4 7 1 1 3)
G range( 4 7 1 1 3)
H range( 7 1 1 3)
How can do this in awk?
I could do the same in perl using:
my $set_size = #values;
for ( my $i = 0 ; $i < $set_size ; $i++ ) {
my $min = $i - 4;
if ( $min < 0 ) { $min = 0; }
my $max = $i + 4;
if ( $max > ( $set_size - 1 ) ) { $max = $set_size - 1; }
my $min_val = $values[$min];
my $max_val = $values[$min];
for ( my $j = $min ; $j <= $max ; $j++ ) {
if ( $values[$j] <= $min_val ) { $min_val = $values[$j]; }
if ( $values[$j] >= $max_val ) { $max_val = $values[$j]; }
}
my $range = $max_val - $min_val;
printf "$points[$i] %.15f\n", $range;
}

Answer 1

idk exactly what I want to calculate the range (max-min) of 3 field radius up and down. means but to get the output you posted from the input you posted would be:
$ cat tst.awk
{
keys[NR] = $1
values[NR] = $2
}
END {
range = 3
for (i=1; i<=NR; i++) {
min = ( (i - range) >= 1 ? i - range : 1 )
max = ( (i + range) <= NR ? i + range : NR )
printf "%s range(", keys[i]
for (j=min; j<=max; j++) {
printf "%s%s", values[j], (j<max ? " " : ")\n")
}
}
}
$
$ awk -f tst.awk file
A range(1 3 5 4)
B range(1 3 5 4 7)
C range(1 3 5 4 7 1)
D range(1 3 5 4 7 1 1)
E range(3 5 4 7 1 1 3)
F range(5 4 7 1 1 3)
G range(4 7 1 1 3)
H range(7 1 1 3)

Answer 2

Your sample perl doesn't print out your sample output, it seems to do something different... so here's how I'd do it in perl:
#!/usr/bin/perl
use warnings;
use strict;
use feature qw/say/;
use List::Util qw/min max/;
my (#col1, #col2);
while (<>) {
chomp;
my ($v1, $v2) = split;
push #col1, $v1;
push #col2, $v2;
}
my #prefix;
for my $i (0 .. $#col1) {
my #range = #col2[max($i - 3, 0) .. min($i + 3, $#col2)];
push #prefix, ' ' if $i > 3;
unshift #range, #prefix;
say "$col1[$i] range(#range)"
}
running it:
$ perl range.pl input.txt
A range(1 3 5 4)
B range(1 3 5 4 7)
C range(1 3 5 4 7 1)
D range(1 3 5 4 7 1 1)
E range( 3 5 4 7 1 1 3)
F range( 5 4 7 1 1 3)
G range( 4 7 1 1 3)
H range( 7 1 1 3)
The formatting will break if any of the numbers are greater than 9, though.

Answer 3

Since you tagged tcl:
#!/usr/bin/env tclsh
lassign $argv file size
set fh [open $file r]
# assume exactly 2 space-separated words per line
set data [dict create {*}[split [read -nonewline $fh]]]
close $fh
set len [dict size $data]
set values [dict values $data]
set i 0
dict for {key _} $data {
set first [expr {max($i - $size, 0)}]
set last [expr {min($i + $size, $len)}]
puts [format "%s range(%s%s)" \
$key \
[string repeat " " $first] \
[lrange $values $first $last] \
]
incr i
}
outputs
A range(1 3 5 4)
B range(1 3 5 4 7)
C range(1 3 5 4 7 1)
D range(1 3 5 4 7 1 1)
E range( 3 5 4 7 1 1 3)
F range( 5 4 7 1 1 3)
G range( 4 7 1 1 3)
H range( 7 1 1 3)

Answer 4

TXR at the shell prompt:
bash$ txr -c '#(collect)
#label #num
#(end)
#(bind rng #[window-map 3 nil (opip list (remq nil) (mapcar toint)) num])
#(output)
# (repeat)
#label range(#rng) -> #(find-min rng)..#(find-max rng)
# (end)
#(end)'
A 1
B 3
C 5
D 4
E 7
F 1
G 1
H 3
[Ctrl-D][Enter]
A range(1 3 5 4) -> 1..5
B range(1 3 5 4 7) -> 1..7
C range(1 3 5 4 7 1) -> 1..7
D range(1 3 5 4 7 1 1) -> 1..7
E range(3 5 4 7 1 1 3) -> 1..7
F range(5 4 7 1 1 3) -> 1..7
G range(4 7 1 1 3) -> 1..7
H range(7 1 1 3) -> 1..7