Thanks, Taco.
     As we talked, since the large channel number (16384) of each SWARM correlator chunks is not limited by either CASA or Miriad,
we have an option in Miriad to let users decide if they need to reduce the spectral resolution while they make CASA output for spectral line imaging.
For wideband continuum imaging, we will take the vector average 's49=s50:1000:15000:10' to reduce the number of channels (or data size).

     Jun-Hui
    

On Thu, Sep 11, 2014 at 10:54 AM, Young, Ken  wrote:

    Dear SMA CASA Enthusiast,

from:	 Young, Ken 
to:	 Mark Gurwell ,
 Eric Keto ,
 Chunhua Qi ,
 Jun-Hui Zhao ,
 Qizhou Zhang 
date:	 Thu, Sep 11, 2014 at 10:54 AM
subject:	 Rebinning SWARM data with sma2casa.py
Dear SMA CASA Enthusiast,

   We will need a way to reduce the resolution of SWARM data when we process it in CASA.   Charlie has already implemented a way to do this in MIR.   I plan to implement the following command-line option to sma2casa.py.   sma2casa.py will accept a variable number of parameters of the form

s{n}=s{m}:{i}-{j}:{q}

where n is a new synthetic chunk number, m is the chunk number from either the ASIC or SWARM, i is the starting channel of chunk m to use, j is the ending channel, and q is the number of channels in m to vector average to produce a channel in n.   So, for example,

s49=s50:1000:15000:10

would produce a chunk (spw) in CASA called spw 49 which would contain the SWARM data from chunk 50, starting with channel 1000 and ending with channel 1500.   10 SWARM channels would be averaged to produce each s49 channel.

Does that syntax look OK to you?   Can you think of a better one?

Thanks,

Taco