[SOLVED] How to delete header (first row) and first column from multiple CSV files

Haba2015 · 03-23-2015, 09:47 AM

I have several of csv files, I want to remove the first row (header) and first column as well. This is my code, but it overwrites the original files with no content inside:

for x in *.csv;
do
sed '1d' $x | cut -d, -f 1 > "$x";

done

Help would be highly appreciated.

linosaurusroot · 03-23-2015, 10:21 AM

You could use a temporary filename if you can think of one that's definitely not in use.

[code]
sed '1d' $x | cut -d, -f 1 > replace_$x && mv replace_$x $x
[//code]

Haba2015 · 03-23-2015, 01:49 PM

Hi linosaurusroot,

the code: for x in *.csv; do sed '1d' $x | cut -d, -f 1 > replace_$x && mv replace_$x $x ; done is deleting the files and only retain the column 1 , I am intending to cut

Haba2015 · 03-23-2015, 04:03 PM

Thank you: This works:

for x in *.csv; do sed '1d' $x| awk 'BEGIN{FS=OFS=","}{$1="";sub(",","")}1' > rippe_$x && mv rippe_$x $x ; done

linosaurusroot · 03-24-2015, 05:51 AM

I used your code from the earlier post

Code:

cut -d, -f 1

instead of

Code:

cut -d, -f 2-

pan64 · 03-24-2015, 06:03 AM

if awk:

Code:

awk ' BEGIN{FS=OFS=","}NR>1{$1="";sub(",","");print}'

and no sed required

syg00 · 03-24-2015, 06:22 AM

Or even

Code:

awk ' BEGIN{FS=OFS=","}NR>1{sub(/[^,]+,/,"");print}'

grail · 03-24-2015, 07:46 AM

I see no issue with sed either:

Code:

sed -i -e '1d' -e 's/[^,]*,//' *.csv

If you append something to 'i' option it will also make a backup, eg -i.bak

syg00 · 03-24-2015, 07:56 AM

Not like you grail to use two "-e" when one would suffice ....