mjhjmtu

I need to process a file like below in the bash script:

input.txt:

host1 53
host1 123
host2 0
host1 222
host3 1
host1 85
host1 25
host1 13
host3 8
host2 90

I need to get in the results only one line for each host based on maximum
value in column 2:

output.txt:

host1 222
host2 90
host3 8

Any ideas?

With GNU sort or compatible:

<input.txt sort -k2rn | sort -sbuk1,1 >output.txt

The shortest one with GNU datamash:

datamash -sW -g1 max 2 <file

• 
  -s - sort the input before grouping


• 
  -W - use whitespace (one or more spaces and/or tabs) for field delimiters


• 
  -g 1 - group via 1st field


• 
  max 2 - numeric operation to get the maximum value of the 2nd field per group

The output:

host1 222
host2 90
host3 8

Here's the awk one-liner:

$ awk ' END for (k in a) print k, a[k] ' < input.txt
host1 222
host2 90
host3 8

$1 and $2 are the first two fields of the line, associative array a[k] holds the current maximum for key k. With the test against a[$1] being empty, it should deal with negative numbers correctly.