NRAO Home  >  Green Bank  |  Wiki Topic:    GB > Data > IDLDocumentation > GBTIDLNsaveRecipe

Changes | Index | Contents | Search | Statistics | Jump to Topic

Examples of Using the nsave Feature in GBTIDL

The nsave feature allows a user to store data into a keep file and attach an identifier number (the nsave value) to that entry. Data can then be stored, overwritten, and retrieved according to the nsave value. The utility essentially gives the user access to an unlimited number of storage slots on disk, somewhat like the 16 data containers kept in memory.

This page gives some examples of how the nsave feature can be used.

The following GBTIDL procedures are relevant to the nsave features. Click the link to go to the reference manual entry.

The following sequence shows how to store and retrieve a spectrum using the nsave feature:

   fileout,'mynsave.fits'     ; open a file for writing
   getrec, 10                 ; get some data
   nsave, 101                 ; store it to the keep file
   scale, 5                   ; perform some operations
   nsave, 102                 ; store it with a different nsave value
   nget, 101                  ; retrieve the spectrum at nsave=101

The next example shows a more sophisticated nsave example. Here each nsave entry stores a calibrated integration from a scan. The example demonstrates how the nsave values can be overwritten and each nsave entry has an attached meaning to the data, for example nsave=1002 is the data associated with scan=100, int=2. As the calibration of this integratin is refined, the spectra are simply stored back into that nsave slot.

   fileout,'mynsave.fits'        ; open a file for writing
   for i=0,5 do begin $          ; store each integration
       getnod, 100, int=i & $
       nsave, 1000+i & $
   endfor

   nget, 1002                    ; retrieve one of the entries
   bias, 0.1                     ; do some work on it
   nsave, 1002                   ; and reinsert it into the file

   for i=0,5 do begin $          ; Now execute a loop to average all
       nget, 1000+i & $          ; the data including the processed
       accum & $                 ; integration.  This loop could be
   endfor                        ; made into a separate procedure.
   ave
   

-- JimBraatz - 06 Jun 2005