AWK每列的RANK值 [英] RANK values per column with AWK

问题描述

AWK中是否有一个RANK函数，可在单个文本文件中对每一列的值进行RANK排序(较高的值排在最前面)?期望这样:

Is there a RANK function in AWK that RANKs values per column in a single text file (higher value ranks on top)? Expecting this:

Original       Results
A  B  C     Result_A  Result_B Result_C
8  5  4     1          2          1
4  7  3     2          1          2
2  3  4     3          3          1

我正在使用此AWK脚本，但仅对单个列进行排名.谁能引导我正确的方向?非常感谢.

I'm using this AWK script but only Ranks a single column. Can anyone guide me to the right direction? Very appreciated.

awk ' {
if(val!=$2){ rank++; }
printf("%s\t%s\t%s\n",rank,$1,$2)
val=$2
} ' <(sort -k2 -nr Original.txt) > Results.txt

推荐答案

您可以尝试使用 gawk 和 asorti 函数

You can try with gawk and asorti function Array-Sorting-Functions

awk 'BEGIN{OFS="\t"}
    NR==1{
        for(i=1; i<=NF; ++i) printf "Result_%s%s", $i, i==NF?ORS:OFS 
        next
    }
    { for(i=1; i<=NF; ++i) M[i, NR-1]=$i }
    END{
        for(i=1; i<=NF; ++i){
            delete d;
            for(j=1; j<=NR-1; ++j) d[M[i,j]]
            n = asorti(d, idx)
            for(j=1; j<=n; ++j) rank[i, idx[j]] = n-j+1
        }
        for(j=1; j<=NR-1; ++j)
            for(i=1; i<=NF; ++i)
                printf "%s%s", rank[i, M[i,j]], i==NF?ORS:OFS
    }
' original

你得到

Result_A    Result_B    Result_C
1   2   1
2   1   2
3   3   1

这篇关于AWK每列的RANK值的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！