Multiple file combination and data evaluation

[rklad77]

Hello!!
respected gurus.

I have this problem of combining multiple files with 2 columns each and then finding out the closest match to the reference column.
I will try to explain it more explicitly:-

I have got first file called "test" with number of rows and only two columns separated by comma. The second column in this file we can call as "reference column".

file "test" reads

5,11
10,22
15,33
..,.. and so on

After doing number of experiments I have got number of files
i called "Run1", "Run2", etc. each with its first column
same as "test".

example
file "Run1" reads
5,10
10,24
15,35
..,.. etc

first I need to combine the "test" file and "Run1 --RunN" files. so that it has common first column and rest of columns from "test" and "Run.." files.

After this file is formed i need to find the sum of squares of differences between the "reference column" and "Run.." columns i.e

if i take first three rows of the "test" file and "Run1" file mentioned above

i get combined file as:

5,11,10,
10,22,24
15,33,25
------------

and i need:-

A1= sq(11-10)+ sq(22-24) + sq(33-35)

where:-
sq= square
A1= sum of square of differences for reference column and
run1 column.

the programme should give these answers for all the run files "Run1", "Run2" etc.

I hope i am clear.

Thanks in advance

 

[CaKiwi]

Try

awk -f rklad77.awk test run1 run2 ...

# ------ rklad77.awk ------
BEGIN {FS=","}
{
if (FNR==1) n1++
a[n1,$1] = $2
if (n1==1) b[++n2] = $1
}
END{
for (j=1;j<=n2;j++) {
printf b[j] &quot;,&quot;
for (k=1;k<=n1;k++) {
printf a[k,b[j]]
if (k < n1) printf &quot;,&quot;
}
print &quot;&quot;
}
}

CaKiwi

&quot;I love mankind, it's people I can't stand&quot; - Linus Van Pelt

 

[Ygor]

Create a script called prog.awk....

#!/usr/bin/awk -f
BEGIN { FS=&quot;,&quot;;
while (getline<&quot;test&quot
test[$1]=$2}
FNR==1 {name[++n]=FILENAME}
{klud[FILENAME,$1]=$2}
END {for (i in name) {sum=0;
for (j in test)
sum+=(test[j]-klud[name,j])^2;
print name &quot;:&quot;, sum } }

Run it like this....

prog.awk Run* | sort -n +1

Tested....

Run1: 9
Run2: 53

 

[rklad77]

Thanks both the programmes worked.
Its of great help.