Hi,

I have three large data files that need alteration. Each consists of -9999 (null data) entries and 5 or 6 digit numeric entries (104625, 99841, etc). I would like to divide the numeric entries by 100000 while allowing the null data entries to remain intact. There is also six rows of "header information" that needs to be skipped over as the alteration is performed.

A small portion of one file:

ncols 1484
nrows 1680
xllcorner -117.231250
yllcorner 41.98958333
cellsize 0.00416667
nodata_value -9999
-9999 -9999 -9999 -9999 -9999 -9999 103456 102321 109884 102223
-9999 -9999 -9999 -9999 -9999 -9999 99667 94557 103445 103224

and so on...

i would like to have:

ncols 1484
nrows 1680
xllcorner -117.231250
yllcorner 41.98958333
cellsize 0.00416667
nodata_value -9999
-9999 -9999 -9999 -9999 -9999 -9999 1.03456 1.02321 1.09884 1.02223
-9999 -9999 -9999 -9999 -9999 -9999 0.99667 0.94557 1.03445 1.03224

Thank you in advance for your help!

Hi malinut89,
try this :

#!/bin/ksh

I="mydata.org" # your inputfile
O="mydata.new" # your outputfile
> $O

cat $I | awk ' BEGIN { FS=" "; i=0}
{
#### maximum values per row
max = 10

#### read the nodata_value
if ( $1 == "nodata_value" )
{
ndv = $2
}
i=i+1
#### write the the first six rows as header to new file
if (i < 7)
{
print $0 > "mydata.new"
}
else
{
x=0
while (x != max)
{
x=x+1
#### null data entries to remain intact
if ( $x != ndv )
{
$x = $x / 100000
}
}
x=0
while (x != max)
{
x=x+1
printf ("%s " ,$x) > "mydata.new"
}
printf ( "\n") > "mydata.new"

}
}'

nawk -f mali.awk myFiles*

here's mali.awk: