Question

(BASH) Say I have an array of strings with about 100 numbers ranging in length from 3-6 that represent some sort of ID. Within the array of strings, there are some duplicates numbers. What I am trying to do is represent only the top 12 numbers in the form (ID, count occurence) where its ranked by the count occurrences. So the first number would not be the ID with the most digits but it would be the one with the most occurrences.  Is there a way to do this with AWK , begin? or how could i approach this problem?

Answer 1

Solution:

• Convert given array of strings to a string

For example, for array

declare -a arr=("ABS" "ABC" "SAF" "ABS" "ABC" "SAF" "ABS" "ABC" "SAF" "SAF")

string is str = "ABS ABC SAF ABS ABC SAF ABS ABC SAF SAF"

• Use below command that count occurence and then sort and print top 12 results.

Command: echo $str | tr '[:space:]' '[\n*]' | sort | uniq -c | sort -bnr | head -12

where,

• echo: print the result
• tr: just replaces spaces in string with newlines
• sort: to prepare input for uniq
• uniq -c: to count occurences
• sort -bnr: sorts in descending order
• head -12: prints top 12 entries

Complete bash code:

str=""
## declare an array variable
declare -a arr=("ABS" "ABC" "SAF" "ABS" "ABC" "SAD" "ABS" "ABC" "SAF" "SAF" "ABS" "ABC" "SAG" "ABD" "DBC" "CAF" "CAF" "ABD" "SDF" "SSF" "ABS" "ABC" "SAF" "ABS" "ABC" "SRF" "ACS" "ABD" "SRF" "SEF")

## loop through the above array and convert to string str
for i in "${arr[@]}"
do
   str="${str} $i"
done

## Find top 12 id with highest count occurence

echo $str | tr '[:space:]' '[\n*]' | sort | uniq -c | sort -bnr | head -12 | awk '{print $2","$1}'

Output:

ABS,6
ABC,6
SAF,4
ABD,3
SRF,2
CAF,2
SSF,1
SEF,1
SDF,1
SAG,1
SAD,1
DBC,1

Attached screenshot of code:

PS: Let me know if you have any doubt.