On 12 April 2012 23:16, Ryo Furue
<furue@xxxxxxxxxx> wrote:
Hi Szymon and Ansley,
> The FILE command (an alias for SET DATA/EZ) only initializes the
> dataset. It checks that it can open the file, and stores the
> definitions of the variables and grids that you give in the FILE
> command. It does not read any data or do anything to see what's in
> the file. This means that each FILE command is just redefining the
> contents of the dataset.
>
> My suggestion would be to read each variable and save it to a netCDF
> file. Then you would have them together in a dataset that's easy to
Another idea is to give multiple names to the datafile using
the symbolic-link mechanism. At the end of this message,
I show a complete example.
Regards,
Ryo
--------------
linux-prompt$ cat tmp.dat #-- contains one 1D variable on each line.
1 2 3 2 1 2 3 2 1
-3 -2 -1 -2 -3 -2 -1 -2 -3
linux-prompt$ ln -s tmp.dat tmp2.dat
linux-prompt$ ferret
. . . .
yes? define axis/x=1:9:1 xax
yes? define grid/x=xax mygrid
yes? file/var=var1/columns=9/grid=mygrid tmp.dat
yes? file/var=var2/columns=9/grid=mygrid/skip=1 tmp2.dat
yes? show data
currently SET data sets:
1> ./tmp.dat
name title I J K L
VAR1 VAR1 1:9 ... ... ...
2> ./tmp2.dat (default)
name title I J K L
VAR2 VAR2 1:9 ... ... ...
yes? plot var1[d=1], var2[d=2]