Disk Based Processing

Normally scantables exist entirely in memory during an ASAP session. This has the advantage of speed, but causes limits on the size of the dataset which can be loaded. ASAP can use ``disk based'' scan tables which cache the bulk of the scantable on disk and require significantly less memory usage. This should be used for all MOPS data!

To use disk based tables you either need to change the default in your .asaprc file, e.g.

   scantable.storage          : disk

or use set the ``rc'' value while running asap to change this on-the-fly. E.g.

  ASAP>data = scantable('data.rpf')     # Loaded using disk based table
  ASAP>rc('scantable',storage='memory') # Memory tables will be used now

Changing the ``rc'' value affects the next time the scantable constructor is called.

NOTE: Currently a bug in ipython means temporary files are not cleaned up properly when you exit ASAP. If you use disk based scan tables your directory will be left with 'tmpXXXXX_X' directories. These can be safely removed if ASAP is not running.

Malte Marquarding 2007-08-16