1

I am trying to find everywhere my data has a 90 in column 2 and two lines above change the value of column 2. For example in my data below, if I see 90 at line 11 I want to change my column 2 value at line 9 from 11 to 5. I have a predetermined set of values I want to change the number to; the values will always be 10,11,12,30,31,32 to 1,2,3,4,5,6 respectably.

My data

 #      Type    Response        Acc     RT      Offset    
   1      70  0    0   0.0000 57850
   2      31  0    0   0.0000 59371
   3      41  0    0   0.0000 60909
   4      70  0    0   0.0000 61478
   5      31  0    0   0.0000 62999 
   6      41  0    0   0.0000 64537
   8      70  0    0   0.0000 65106
   9      11  0    0   0.0000 66627
  10      21  0    0   0.0000 68165
  11      90  0    0   0.0000 68700
  12      31  0    0   0.0000 70221

What I want

 #      Type    Response        Acc     RT      Offset    
   1      70  0    0   0.0000 57850
   2      31  0    0   0.0000 59371
   3      41  0    0   0.0000 60909
   4      70  0    0   0.0000 61478
   5      31  0    0   0.0000 62999 
   6      41  0    0   0.0000 64537
   8      70  0    0   0.0000 65106
   9       5  0    0   0.0000 66627
  10      21  0    0   0.0000 68165
  11      90  0    0   0.0000 68700
  12      31  0    0   0.0000 70221

I have been trying to store the previous line and use that as a reference but I can only go back one line, and I need to go back two. Thank you for your help.

4

2 回答 2

1

This should work:

function pra(a) {
  for(e in a) {
    printf "%s ", a[e];
  }
  print ""; 
}
BEGIN {
  vals[10] = 1;
  vals[11] = 2;
  vals[12] = 3;
  vals[30] = 4;
  vals[31] = 5;
  vals[32] = 6;
}
NR == 1 { split($0, a, " ") }
NR == 2 { split($0, b, " ") }
NR > 2 { 
  if($2 == "90") {
    a[2] = vals[a[2]];
  }
  pra(a);
  al = 0;
  for(i in a) al++;
  for(i = 1; i <= al; i++) {
    a[i] = b[i];
  }
  split($0, b, " ");
}
END {
  pra(a);
  pra(b);
}

The rundown of how this works: * BEGING block - assign the translation values to vals * NR == 1 and NR == 2 - remember the first two lines into split arrays a and b * NR > 2 - for all lines after the first two * If the second column has value 90, change it using the translation array * Move elements of array b to a and split the current line into b * END block - print a and b, which are basically last two lines

Sample run:

$ cat inp && awk -f mkt.awk inp 
 #      Type    Response        Acc     RT      Offset    
   1      70  0    0   0.0000 57850
   2      31  0    0   0.0000 59371
   3      41  0    0   0.0000 60909
   4      70  0    0   0.0000 61478
   5      31  0    0   0.0000 62999 
   6      41  0    0   0.0000 64537
   8      70  0    0   0.0000 65106
   9      11  0    0   0.0000 66627
  10      21  0    0   0.0000 68165
  11      90  0    0   0.0000 68700
  12      31  0    0   0.0000 70221

# Type Response Acc RT Offset 
1 70 0 0 0.0000 57850 
2 31 0 0 0.0000 59371 
3 41 0 0 0.0000 60909 
4 70 0 0 0.0000 61478 
5 31 0 0 0.0000 62999 
6 41 0 0 0.0000 64537 
8 70 0 0 0.0000 65106 
9 2 0 0 0.0000 66627 
10 21 0 0 0.0000 68165 
11 90 0 0 0.0000 68700 
12 31 0 0 0.0000 70221

You can do something like this:

function pra(a) {
  printf "%4d%8d%3d%5d%9.4f%6d\n", a[1], a[2], a[3], a[4], a[5], a[6]
}
BEGIN {
  vals[10] = 1;
  vals[11] = 2;
  vals[12] = 3;
  vals[30] = 4;
  vals[31] = 5;
  vals[32] = 6;
}
NR == 1 { print }
NR == 2 { split($0, a, " ") }
NR == 3 { split($0, b, " ") }
NR > 4 {
  if($2 == "90") {
    a[2] = vals[a[2]];
  }
  pra(a);
  for(i = 1; i <= 6; i++) {
    a[i] = b[i];
  }
  split($0, b, " ");
}
END {
  pra(a);
  pra(b);
}

To make it work for this specific case that includes formatting. Sample run:

$ cat inp && awk -f mkt.awk inp 
 #      Type    Response        Acc     RT      Offset    
   1      70  0    0   0.0000 57850
   2      31  0    0   0.0000 59371
   3      41  0    0   0.0000 60909
   4      70  0    0   0.0000 61478
   5      31  0    0   0.0000 62999 
   6      41  0    0   0.0000 64537
   8      70  0    0   0.0000 65106
   9      11  0    0   0.0000 66627
  10      21  0    0   0.0000 68165
  11      90  0    0   0.0000 68700
  12      31  0    0   0.0000 70221 
 #      Type    Response        Acc     RT      Offset    
   1      70  0    0   0.0000 57850
   2      31  0    0   0.0000 59371
   4      70  0    0   0.0000 61478
   5      31  0    0   0.0000 62999
   6      41  0    0   0.0000 64537
   8      70  0    0   0.0000 65106
   9       2  0    0   0.0000 66627
  10      21  0    0   0.0000 68165
  11      90  0    0   0.0000 68700
  12      31  0    0   0.0000 70221
于 2012-03-18T02:31:45.230 回答
0

This version maintains your original formatting 

awk 'BEGIN{ new[" 1"]="10"; new[" 2"]="11"; new[" 3"]="12"
            new[" 4"]="30"; new[" 5"]="31"; new[" 6"]="32" }
     { line[-2]=line[-1]; line[-1]=line[0]; line[0]=$0 } 
     $2==90 { if( match( line[-2], /^ *[0-9]+ +[1-6] / ) ) { 
                  old=substr( line[-2], RLENGTH-2,2 )
                  line[-2]=substr( line[-2], 1, RLENGTH-3 ) new[old] \
                           substr( line[-2], RLENGTH ) } }
     NR>2 { printf("%s\n",line[-2]) } 
     END { printf("%s\n%s\n",line[-1],line[0]) }' file.in
于 2012-03-18T05:18:15.213 回答