Run query in Linux for selecting CSV'S - linux

Question

In the Linux:
there are many .csvs' in the folder, I have to select those csv's file having column name {'PREDICT' = 646}.
check this link:
https://prnt.sc/gone85
what kind of query works?

Answer 1

Providing test data which was unprovided ):
$ cat > file1
ACTUAL PREDICT
1 2
3 646
$ cat > file2
ACTUAL PREDICT
1 2
3 666
Then some GNU awk (nextfile) to select those csv's file having column name {'PREDICT' = 646} or where there is column PREDICT with a value 646:
$ awk 'FNR==1{for(i=1;i<=NF;i++)if($i=="PREDICT")p=i}$p==646{print FILENAME;nextfile}' file1 file2
file1
Explained:
awk '
FNR==1 { # get the column number of PREDICT column for each file
for(i=1;i<=NF;i++)
if($i=="PREDICT")
p=i # set it to p
}
$p==646 { # if p==646, we have a match
print FILENAME # print the filename
nextfile # and move on to the next file
}' file1 file2 # all the candicate files

Answer 2

gnu awk solution without loop:
$ cat tst.awk
BEGIN{FS=","}
FNR==1 && s=substr($0,1,index($0,"PREDICT")) { # look for index of PREDICT
i=sub(/,/, "", s) + 1 # and count nr of times you
# can replace "," in preceding
# substring
}
s && $i==646 { print FILENAME; nextfile }
some input:
$ cat file1.csv
ACTUAL,PREDICT,COUNTRY,REGION,DIVISION,PRODUCTTYPE,PRODUCT,QUARTER,YEAR,MONTH
925,850,CANADA,EAST,EDUCATION,FURNITURE,SOFA,1,1993,12054
925,533,CANADA,EAST,EDUCATION,FURNITURE,SOFA,1,1993,12054
925,646,CANADA,EAST,EDUCATION,FURNITURE,SOFA,1,1993,12054
$ cat file2.csv
ACTUAL,PREDICT,COUNTRY,REGION,DIVISION,PRODUCTTYPE,PRODUCT,QUARTER,YEAR,MONTH
925,850,CANADA,EAST,EDUCATION,FURNITURE,SOFA,1,1993,12054
925,533,CANADA,EAST,EDUCATION,FURNITURE,SOFA,1,1993,12054
925,111,CANADA,EAST,EDUCATION,FURNITURE,SOFA,1,1993,12054
and:
$ cp file1.csv file3.csv
gives:
$ awk -f tst.awk *.csv
file1.csv
file3.csv
Or use a one-liner:
$ awk -F, 'FNR==1 && s=substr($0,1,index($0,"PREDICT")) {i=sub(/,/, "", s) + 1}s && $i==646 { print FILENAME; nextfile }' *.csv
file1.csv
file3.csv