Sorting unique by column

Shell Programming and Scripting
1

I am trying to sort, do uniq by 1st column and report this 4 columns tab delimiter table , eg

chr10:112174128 rs2255141       2E-10   Cholesterol, total
chr10:112174128 rs2255141       7E-16   LDL
chr10:17218291  rs10904908      3E-11   HDL Cholesterol
chr10:17218291  rs970548        8E-9    TG
chr1:109275684  rs629301        2E-170  Cholesterol, totat

I want an output this this

chr10:112174128 rs2255141       2E-10/7E-16   Cholesterol, total; LDL
chr10:17218291  rs10904908/rs970548      3E-11/8E-9   HDL Cholesterol; TG
chr1:109275684  rs629301        2E-170  Cholesterol, total

Can you help ?

2

Any attempts from your side?

3

Yes,

awk 'BEGIN{printf "\t"}{A[$1] = A[$1] ?  A[$1] OFS $4 : $4}FNR==1{printf FILENAME OFS}END{printf RS; for(i in A)print i, A}' OFS='\t' myfile.txt | sort
4

Try

awk     '       {F2[$1]=F2[$1]"/"$2
                 F3[$1]=F3[$1]"/"$3
                 F4[$1]=F4[$1]";"$4}
         END    {for (S1 in F2)
                         print  S1,
                                substr(F2[S1],2),
                                substr(F3[S1],2),
                                substr(F4[S1],2)}
        ' FS="\t" OFS="\t" file | sort
chr10:112174128    rs2255141/rs2255141    2E-10/7E-16    Cholesterol, total;LDL
chr10:17218291     rs10904908/rs970548    3E-11/8E-9     HDL Cholesterol;TG
chr1:109275684     rs629301               2E-170         Cholesterol, totat
1 Like
5

If we sort first can save using memory like this:

sort infile | awk '
F1 && F1!=$1 {print F1,F2,F3,F4; F2=F3=F4=x}
{ F1=$1
  F2=(F2?F2"/":x)$2
  F3=(F3?F3"/":x)$3
  F4=(F4?F4";":x)$4 }
END {print F1,F2,F3,F4}' FS='\t' OFS='\t'
chr1:109275684  rs629301                2E-170          Cholesterol, totat
chr10:112174128 rs2255141/rs2255141     2E-10/7E-16     Cholesterol, total;LDL
chr10:17218291  rs10904908/rs970548     3E-11/8E-9      HDL Cholesterol;TG