Moving Data from a Column to a Row/String

[danhodges99]

Hi all,

I have a file with several columns of data, eg:

A B C D
--- --- --- ---
1 2 5 1
2 2 2 2
8 4 4 4
4 2 3 4
10 9 4 4
9 7 1 2

I need to get the values from, say, column B and place them into a string separated by a semicolon, eg:

2;2;4;2;9;7

Does anyone have any ideas how to do this? Any help would be massively appreciated! Thanks.

 

[vgersh99]

something like that:

nawk -v colID="B" -f getCol.awk text.txt
nawk -v colID="D" -f getCol.awk text.txt

#----------------------- getCol.awk
BEGIN {
OFS=";"

if (colID=="&quot colID="A";

FLDheader="1"
FLDskip="2"
}

NR == FLDheader {
for(i=1; i <= NF; i++)
if ($i == colID) {
colNum=i
break;
}
if (colNum == &quot;0&quot {
printf(&quot;Warning: couldn't find column [%s]\n&quot;, colID) | &quot;cat 1>&2&quot;
exit 1;
}
next;
}

NR != FLDskip {
row = (length(row)) ? row OFS $colNum : $colNum;
}

END {
if (colNum != 0)
print row;
}


vlad
+----------------------------+
| #include<disclaimer.h> |
+----------------------------+

 

[vgersh99]

or better yet - to skip 'empty' lines. This produced an undesired extra ';' you're posting about in your other thread in this forum.

BEGIN {
OFS=&quot;;&quot;

if (colID==&quot;&quot colID=&quot;A&quot;;
colNum=&quot;0&quot;

FLDheader=&quot;1&quot;
FLDskip=&quot;2&quot;
}

NR == FLDheader {
for(i=1; i <= NF; i++)
if ($i == colID) {
colNum=i
break;
}
if (colNum == &quot;0&quot {
printf(&quot;Warning: couldn't find column [%s]\n&quot;, colID) | &quot;cat 1>&2&quot;
exit 1;
}
next;
}

NR != FLDskip && NF >= colNum {
row = (length(row)) ? row OFS $colNum : $colNum;
}

END {
if (colNum != 0)
print row;
}




vlad
+----------------------------+
| #include<disclaimer.h> |
+----------------------------+