Re: [ferret_users] Looping over multile HDF files.

[Ansley Manke <ansley.b.manke@xxxxxxxx>]

Hello C. Shaji,
What I would do is to write a script that does all the operations you 
want to do on each file.  Then call this from another script, sending 
the month number, or perhaps the file path as an argument.  You could 
also send in arguments with variable names, directory paths, and so on, 
if those things are not going to be the same

 ! script write_file.jnl
 ! open a monthly HDF file and re-write it as a binary file
 ! Argument: month number

 ! symbol nmon with month, e.g. 01, or 11
 define symbol nmon = $1
 if `($nmon) LT 10` THEN define symbol nmon = 0$1

 use
"http://data.nodc.noaa.gov/cgi-bin/nph-dods/pathfinder/Version5.0_Climatologies/Monthly/Combined/month($nmon)_combined.hdf"

 !Shows the dataset variables.

 show data
 show grid

 SHOW GRID CLIM_SST

 ! To get the axis information
 SHOW AXIS LONGITUDE
 SHOW AXIS LATITUDE

 SET MEMORY/SIZ=416

 
list/format=stream/file="/work/data/sst/pf/data/monthly/bin/month($nmon)_combined.bin" 
clim_sst

 let lat = Y[GY=clim_sst]
 let lon = X[GX=clim_sst]

 
list/format=stream/file="/work/data/sst/pf/data/monthly/bin/month($nmon)_combined_lat.grd" 
lat

 
list/format=stream/file="/work/data/sst/pf/data/monthly/bin/month($nmon)_combined_lon.grd" 
lon

  ! End of script write_file.jnl

Then call this script for each month:

 ! script all_months.jnl
 repeat/range=1:12:1/name=m (go write_file `m`)

Or, if for instance you found that some files didn't exist, this could 
also be called without a loop.  It's easy to use an editor to make a 
list of commands:

  go write_file 1
  go write_file 3
  go write_file 4
  go write_file 5
  go write_file 6
  go write_file 9
...

Ansley

Dr.C.Shaji wrote:
> Dear ferret users,
> I am remotely accessing 12 HDF files one by one. Finally I would like to
> write the data as binary in another 12 files. How can I do this using a
> file loop.
>
> 1) For example, for a single file, I use the following commands to achieve
> this.
>
> use
> "http://data.nodc.noaa.gov/cgi-bin/nph-dods/pathfinder/Version5.0_Climatologies/Monthly/Combined/month01_combined.hdf";
> !January.
> !Shows the dataset variables.
>
> show data
> show grid
>
> SHOW GRID CLIM_SST
>
> ! To get the axis information
> SHOW AXIS LONGITUDE
> SHOW AXIS LATITUDE
>
> SET MEMORY/SIZ=416
>
> list/format=stream/file="/work/data/sst/pf/data/monthly/bin/month01_combined.bin"
>  clim_sst !January
>
> let lat = Y[GY=clim_sst]
> let lon = X[GX=clim_sst]
>
> list/format=stream/file="/work/data/sst/pf/data/monthly/bin/month01_combined_lat.grd"
>  lat
>
> list/format=stream/file="/work/data/sst/pf/data/monthly/bin/month01_combined_lon.grd"
>  lon
>
> 2)
> I want to access files such as month01_combined.hdf,
> month02_combined.hdf,....month12_combined.hdf
>
> Finally the binary files month??_combined.bin, month??_combined_lon.grd,
> month??_combined_lon.grd have to be written in another directory.
>
> Can someone suggest me some useful tips to do this in ferret?
>
> Thanks in advance,
> C. Shaji.
>
>
>
>