降序排列

batty123292

我希望把同一个id对应的数据按权重降序排列。

输入数据：

1. 100012,111102   虹      0.00124634
   
2. 100117,127152   书      0.0205799
   
3. 100117,127152   刘胜新  0.951194
   
4. 100117,127152   元      0.0020853
   
5. 100600,141676   画      0.0241507
   
6. 100601,109134   小小    0.00329752
   
7. 100601,109134   爱      0.00282875
   
8. 100611,100228   宝      0.00188535
   
9. 100611,100228   翡翠    0.0125502
   
10. 100611,100228   珍      0.0054723

复制代码

希望输出：
100012,111102   虹：0.00124634
100117,127152   刘胜新：0.951194   书：0.0205799  元：0.0020853
100600,141676   画：0.0241507
100601,109134  小小：0.00329752 爱：0.00282875
100611,100228  翡翠：0.0125502 珍：0.0054723  宝：0.00188535

关阴月飞

回复 1# batty123292

先用sort排好序再交给awk处理

1. [root@master awktest]# cat urfile
   
2. 100012,111102   虹      0.00124634
   
3. 100117,127152   书      0.0205799
   
4. 100117,127152   刘胜新  0.951194
   
5. 100117,127152   元      0.0020853
   
6. 100600,141676   画      0.0241507
   
7. 100601,109134   小小    0.00329752
   
8. 100601,109134   爱      0.00282875
   
9. 100611,100228   宝      0.00188535
   
10. 100611,100228   翡翠    0.0125502
    
11. 100611,100228   珍      0.0054723
    
12. [root@master awktest]# sort -k1n -k3nr ufile |awk '{a[$1]=a[$1]FS$2":"$NF}END{for(i in a)print i,a[i]|"sort"}'
    
13. 100012,111102  虹:0.00124634
    
14. 100117,127152  刘胜新:0.951194 书:0.0205799 元:0.0020853
    
15. 100600,141676  画:0.0241507
    
16. 100601,109134  小小:0.00329752 爱:0.00282875
    
17. 100611,100228  翡翠:0.0125502 珍:0.0054723 宝:0.00188535

复制代码

blackold

try:

sed:

1. sort -sk1,1 -k3nr,3 urfile | sed ':n;N;s/ \+\([0-9.]\+\)/:\1/g;s/^\(\([^ ]\+\) .*\)\n\2 /\1 /;/\n/!tn;P;D'

复制代码

awk:

1. sort -s -k1,1 -k3nr,3 urfile | awk '{if($1==p)printf "\t"$2":"$3;else {if (NR!=1) print "";printf $1"\t"$2":"$3};p=$1}END{print ""}'

复制代码

yestreenstars

awk:

1. sort -k1,1n -k3,3nr i | awk '{if(!a[$1]++)b[++n]=$1;c[$1]=c[$1]?c[$1]" "$2":"$3:$1" "$2":"$3}END{for(i=0;i++<n;)print c[b[i]]}'

复制代码

1. sort -k1,1n -k3,3nr i | awk '{$0=$1" "$2":"$3}t!=$1{printf t?"\n"$0:$0;t=$1;next}{printf " "$2;t=$1}END{print ""}'

复制代码

sed:

1. sort -k1,1n -k3,3nr i | sed -r 's/(\S+)\s+(\S+)\s+(\S+)/\1 \2:\3/;:1;N;s/^(\S+)(.*)\n\1\s+(\S+)\s+(\S+)$/\1\2 \3:\4/;t1;P;D'

复制代码

perl:

1. my %h;
   
2. map { @_ = split; $h{$_[0]}{$_[2]} = $_[1] } <DATA>;
   
3. map { my $k = $_; print ; map { print " $h{$k}{$_}:$_" } sort { $b <=> $a } keys %{$h{$k}}; print $/ } sort keys %h;

复制代码

elu_ligao

1. awk '{a[$1];c[$1,$3]=c[$1]FS$2":"$3;b[$1,$3]=$3}END{for(i in a){printf "%s ",i;split(c[i],sc);l=asorti(b,sb);for(j=l;j>0;--j)printf "%-s ",c[i,b[sb[j]]];print ""}}' idsort | sort | column -t

复制代码

elu_ligao

确实先sort 简单许多了