from:	 Zhao, Jun-Hui 
to:	 Aina Palau 
cc:	 smamiriad@cfa.harvard.edu
date:	 Mon, Oct 27, 2014 at 1:17 PM
subject:	 Re: problems with miriad

Aina,
     Additional option is to consult IDL/MIR developers to ask them which Miriad version is fully compatible
to the Miriad formatted SMA data exported via the IDL/MIR package.
     Good luck.
     JHZ
On Mon, Oct 27, 2014 at 12:36 PM, Zhao, Jun-Hui  wrote:

    Dear Aina,

         Please follow the instruction (http://www.cfa.harvard.edu/sma/miriad/download/) for handling SMA-4GHz-mode data  in Miriad.
    We have made quite effort to help Alice Argon to get rid of the compatibility problems for using IDL/MIR exported Miriad format
    data (http://www.cfa.harvard.edu/sma/miriad/download/userMtg/IDL/). The incompatibility appears to be mainly caused by
    1)  IDL/MIR appears to not strictly follow the Miriad convention.
    2)  The Miriad library that used in IDL/MIR package appears to be out of date (or obsolete).
         You may look at the last entry of the "Test and Support" section provided in http://www.cfa.harvard.edu/sma/miriad/download/
    for  Image scripts for SMA archival calibrated data.
         One of the options may help you out.
     
         Please find a Miriad expert at SMA to help you in properly using Miriad to reduce SMA data. 
         Also please to come to my office (M213, 160 Concord Ave) if you are in Cambridge, we would be happy to help you use
    Miriad for the reduction of SMA data.
         You can also follow the pipeline scripts for calibration/imaging  that are given in the cyber link at SMA Miriad webpage:
    http://www.cfa.harvard.edu/sma/miriad/spec/SMAscripts/  if you are not located in Cambridge. We can do our best to help
    you via email communication.  Please use SMA Miriad beta 1.4.3 or later versions when you execute these scripts. Beware,
    you may also run into incompatibility issues if you do not follow the instruction, since there are many obsolete Miriad versions
    around the world.

         Jun-Hui Zhao

    Astrophysicist at CfA
    SAO, MS 78
    Cambridge, MA 02138
    Tel:  617 496 7895
    Fax: 617 496 7554



    On Mon, Oct 27, 2014 at 11:28 AM, Aina Palau  wrote:

        Hello

        We are having some problems using the MIRIAD beta version 1.4.6 2012-07-08 downloaded from http://www.cfa.harvard.edu/sma/miriad/download/

        When we use the task UVFLAG we get a segmentation fault error:

        Set flags to indicate bad data in visibility file IRAS16293_USB_110824
        Linetype channel; select all channels
        Set flags for the first 3 and the last 3 channels of each window
        Segmentation fault
        finished

        And when we use the task SMAUVSPEC we get an incompatibility error:


        ahernandez@galileo:110824_final_charlie> ./IRAS16293_usb_110824.csh smauvspec
        SmaUvSpec: version beta 27-Jan-2012
        ### Fatal Error [smauvspec]:  Incompatible: Input vis must be SMA data output from SMALOD.
        finished


        Could this be a problem with our software, or with MIRIAD itself? The specifications of the pc where we are executing MIRIAD are the following:

        Linux Fedora 8
        kernel 2.6.26.8-57.fc8
        4 processers Intel Xeon E5405 2.00GHz each
        16 GB RAM


        It seems that Eric Keto also had these problems about 2 years ago (see below), and the answer was that he should use the miriad version of the SMA page. But this is what we have used, and still have the same problems. 

        Any input will be deeply appreciated

        All the best

        Aina




        %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%


        Date: Tue, 4 Sep 2012 16:22:17 -0400 (EDT)
        From: Jun-Hui Zhao 
        To: Eric R. Keto 
        Cc: kyoung@cfa.harvard.edu, rblundell@cfa.harvard.edu, Peter Teuben ,
             melvyn wright 
        Subject: Re: miriad bugs

        Hi Eric,
          I look into more about the version of Miriad /sma/miriad/linux/MIRRC.linux
        that you mentioned. The header says VERSION 4.1.2 but some of the codes
        appear to be quite old. I wonder who is supporting this package
        in which several of my old SMA Miriad codes are mixed with old Miriad
        subroutines. I understand your concerns about Alice's effort. We
        are not able to support those files produced from obsolete
        software. Here is the place that you can find updated versions
        http://www.cfa.harvard.edu/sma/miriad/download/
        Please use the updated software for SMA data reduction
        and also direct SMA users to use the package that we currently
        support.
          Also, we are in the middle of merging the current SMA components
        with the latest version of Miriad as we have discussed in the
        past Miriad Users' meeting (http://www.cfa.harvard.edu/sma/miriad/wbcorrTest/).
        I found that there have been many significant updates in Miriad.
        It takes time to catch up.
          Thanks.
         Jun-Hui

        On Tue, 4 Sep 2012, Eric R. Keto wrote:

>         JH,
>
>         Found a couple of bugs:
>
>         -Eric
>
>          2 Bugs. UVFLAG and UVSPEC
>
>         data set is
>         /sma/SMAusers/keto/pgquasars/100602/cal/100602_030013_PG1613+658_lsb
>
>         FIRST BUG UVFLAG
>         ------------------------------------------------------------------------------
>         cfa0> uvflag vis=$vis edge=8 flagval=flag
>         uvflag: beta 09-Jan-10
>
>         Set flags to indicate bad data in visibility file 100602_030013_PG1613+658_lsb
>         Linetype channel; select all channels
>         Set flags for the first 8 and the last 8 channels of each window
>         Segmentation fault
>         ------------------------------------------------------------------------------
>
>         My speculation:
>
>         The first bug in UVFLAG derives from the MIRIAD distribution. They forgot to have UVFLAG refer

        to the parameter file which specifies the maximum number of windows. UVFLAG is still hard-codes
        the number of windows (chunks).  The MIRIAD version here: /sma/miriad/linux/MIRRC.linux works.

>         SECOND BUG UVSPEC
>
>         ------------------------------------------------------------------------------
>         cfa0> smauvspec vis=$vis interval=9999 device=$dev axis=freq,amp hann=5 nxy=1,1
>         SmaUvSpec: version beta 27-Jan-2012
>         ### Fatal Error [smauvspec]:  Incompatible: Input vis must be SMA data output from SMALOD.
>         ------------------------------------------------------------------------------
>
>         My speculation:
>
>         It's probably not really incompatible. Again, the MIRIAD version here:

        /sma/miriad/linux/MIRRC.linux works.

>         This incompatibility prevents the use of the data that Alice Argon calibrated for the RTDC.

        This calibrated data base is a major effort for the SMA, several years of work by Alice. We have
        to be able to read this data into MIRIAD.


        -- 
        Aina Palau
        Centro de Radioastronomia y Astrofisica
        Universidad Nacional Autonoma de Mexico
        P.O. Box 3-72, 58090, Morelia 
        Michoacan, Mexico
        Phone: +52 56232727-42553