EVLA 6cmWideband Tutorial SN2010FZ: Difference between revisions

From CASA Guides
Jump to navigationJump to search
 
(84 intermediate revisions by 4 users not shown)
Line 1: Line 1:
== Overview ==
#REDIRECT [[EVLA 6cmWideband Tutorial SN2010FZ-CASA4.4]]
This article describes the calibration and imaging of a single-pointing 6cm EVLA wideband continuum dataset on the galaxy NGC2967 (UGC5180) which was the location of the supernova candidate SN2010FZ.  No supernova was detected in this observation, but the galactic continuum emission from this face-on spiral is adequately imaged. The data were taken in RSRO mode, with 1024 MHz of bandwidth in each of two widely spaced basebands (comprised each of 8 128 MHz spectral windows), spanning 4.5 to 7.5 GHz.  We will use wideband imaging techniques in this tutorial.
 
This is a more advanced tutorial, and if you are a relative novice (and <em>particularly</em> for EVLA continuum calibration and imaging), it is <em>strongly</em> recommended that you start with the [[EVLA Continuum Tutorial 3C391]] (at least read it through) before tackling this dataset.  We will not include basic information on CASA processing in this tutorial.
 
From the [[http://casaguides.nrao.edu MainPage]] of this guide you can find helpful information:
* [[What is CASA?]]
* [[Getting Started in CASA]]
* [[CASA Reference Manuals]]
* [[Hints, Tips, & Tricks]]
* [[AIPS-to-CASA Cheat Sheet]]
 
In this tutorial we will be invoking the tasks as function calls.  You can cut and paste these to your casapy session. Occasionally we will be setting Python variables (e.g. as lists for flags) outside the function call so make sure you
set those before running the task command.  Note that when you call a CASA task as a function the task parameters you
do not set in the function call (assuming there is at least one) will be set to their defaults, and will not use values
you set in previous calls or outside the call. See [[Getting_Started_in_CASA#Task_Execution]] for more
on calling tasks and setting parameters in the scripting interface.
 
NOTE: If you find that the figures on the right margin of the browser window overlap the text too much and make reading difficult,
go ahead and widen the browser window.
 
== CASA Versions ==
 
This tutorial was written for the CASA Version 3.2.1 (release r15198 26 May 2011). It has also been verified to work in the stable release Version 3.3.0 (stable r15979 22 Aug 2011).
 
== Obtaining the Data ==
 
The method you should use to get your data depends on whether you are at a NRAO Workshop or doing the tutorial
on your own.
 
=== At the NRAO EVLA Data Reduction Workshop ===
 
If you are attending one of the NRAO EVLA Data Reduction Workshops, then you will be told by the workshop
leaders how to access the data on your workstation.  You will not need to follow the instructions below to
find the data.
 
For these tutorials, you should start from the filled, flagged and averaged MS.
Start with the section below titled "[[#Examining and Flagging your Averaged MS]]".
 
=== On your own ===
 
The scheduling block (SB) processed appears in the EVLA archive under program AS1015 as
<tt>AS1015_sb1658169_1.55388.89474846065</tt>
and was run on 2010-07-11 from 21:28 to 22:28 UT (size 37.74GB).
 
For the purposes of this tutorial, we have provided the raw SDM data (as would be extracted from the archive) as well as measurement sets created by filling the data (with the {{importevla}} task) and upon time-averaging to 10s (after application of the online flags). 
 
To start your tutorial, depending on which dataset you start with, proceed to:
* <em>To start with the raw SDM data:</em> Start with the section below titled "[[#Importing your EVLA data from SDM]]". This is where you would start if you were reducing data from the archive.
* <em>To start with the raw filled MS:</em> Start with the section below titled "[[#Application of Online Flags and Averaging your MS]]".
* <em>To start with the flagged and averaged MS:</em> Start with the section below titled "[[#Examining and Flagging your Averaged MS]]". This is a good place to start if you want to shorten the tutorial.
 
Unless stated otherwise, and where not a direct product of the casapy commands presented here, the data and ancillary masks and images referred to in this tutorial are available on-disk at the AOC in the directory
<tt>/lustre/smyers/AS1015/</tt>.
 
== Importing your EVLA data from SDM ==
 
For the purposes of this tutorial, we assume that the SDM is resident on disk, in this case at the location:
<pre>/lustre/smyers/AS1015/AS1015_sb1658169_1.55388.89474846065</pre> 
Use the actual location of your data when you carry out the commands.
 
The {{listsdm}} task will print out a summary of the scans, fields, spectral windows, and antennas present in your SDM.
 
<source lang="python">
# In CASA
listsdm('/lustre/smyers/AS1015/AS1015_sb1658169_1.55388.89474846065')
</source>
 
In the logger you should see:
<pre>
================================================================================
  SDM File: /lustre/smyers/AS1015/AS1015_sb1658169_1.55388.89474846065
================================================================================
  Observer: Dr. Alicia M. Soderberg
  Facility: EVLA, D-configuration
      Observed from 2010/07/11/21:28:28.41 to 2010/07/11/22:28:17.73 (UTC)
      Total integration time = 3589.32 seconds (1.00 hours)
Scan listing:
  Timerange (UTC)          Scan FldID  FieldName      SpwIDs        Intent(s)
  21:28:28.41 - 21:29:27.40    1    0  J0925+0019      [0, 1]  CALIBRATE_PHASE
  21:29:27.40 - 21:30:57.16    2    0  J0925+0019      [0, 1]  CALIBRATE_PHASE
  21:30:57.16 - 21:32:26.91    3    0  J0925+0019      [0, 1]  CALIBRATE_PHASE
  21:32:26.91 - 21:33:56.67    4    0  J0925+0019      [0, 1]  CALIBRATE_PHASE
  21:33:56.67 - 21:34:56.50    5    0  J0925+0019      [0, 1]  CALIBRATE_PHASE
  21:34:56.50 - 21:35:56.34    6    0  J0925+0019      [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  CALIBRATE_PHASE
  21:35:56.34 - 21:37:26.09    7    0  J0925+0019      [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  CALIBRATE_PHASE
  21:37:26.09 - 21:38:25.93    8    0  J0925+0019      [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  CALIBRATE_PHASE
  21:38:25.93 - 21:39:55.68    9    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  21:39:55.68 - 21:41:25.44  10    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  21:41:25.44 - 21:42:55.19  11    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  21:42:55.19 - 21:44:24.94  12    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  21:44:24.94 - 21:45:54.70  13    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  21:45:54.70 - 21:47:24.45  14    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  21:47:24.45 - 21:47:54.37  15    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  21:47:54.37 - 21:49:24.12  16    0  J0925+0019      [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  CALIBRATE_PHASE
  21:49:24.12 - 21:50:53.88  17    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  21:50:53.88 - 21:52:23.63  18    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  21:52:23.63 - 21:53:53.39  19    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  21:53:53.39 - 21:55:23.14  20    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  21:55:23.14 - 21:56:52.89  21    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  21:56:52.89 - 21:58:22.65  22    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  21:58:22.65 - 21:58:52.57  23    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  21:58:52.57 - 22:00:22.32  24    0  J0925+0019      [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  CALIBRATE_PHASE
  22:00:22.32 - 22:01:52.07  25    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  22:01:52.07 - 22:03:21.83  26    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  22:03:21.83 - 22:04:51.58  27    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  22:04:51.58 - 22:06:21.34  28    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  22:06:21.34 - 22:07:51.09  29    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  22:07:51.09 - 22:09:20.85  30    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  22:09:20.85 - 22:09:50.76  31    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  22:09:50.76 - 22:11:20.52  32    0  J0925+0019      [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  CALIBRATE_PHASE
  22:11:20.52 - 22:12:50.27  33    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  22:12:50.27 - 22:14:20.02  34    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  22:14:20.02 - 22:15:49.78  35    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  22:15:49.78 - 22:17:19.53  36    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  22:17:19.53 - 22:18:49.29  37    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  22:18:49.29 - 22:20:19.04  38    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  22:20:19.04 - 22:20:48.96  39    1  SN2010FZ        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  OBSERVE_TARGET
  22:20:48.96 - 22:22:18.71  40    0  J0925+0019      [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  CALIBRATE_PHASE
  22:22:18.71 - 22:23:48.47  41    2  3C286          [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  CALIBRATE_BANDPASS CALIBRATE_AMPLI
  22:23:48.47 - 22:25:18.22  42    2  3C286          [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  CALIBRATE_BANDPASS CALIBRATE_AMPLI
  22:25:18.22 - 22:26:47.98  43    2  3C286          [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  CALIBRATE_BANDPASS CALIBRATE_AMPLI
  22:26:47.98 - 22:28:17.73  44    2  3C286          [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]  CALIBRATE_BANDPASS CALIBRATE_AMPLI
Spectral window information:
  SpwID  #Chans  Ch0(MHz)  ChWidth(kHz) TotBW(MHz)  Baseband
  0      64      7686.0    2000.0      128.0      BB_4   
  1      64      7836.0    2000.0      128.0      BB_8   
  2      64      4488.0    2000.0      128.0      BB_4   
  3      64      4616.0    2000.0      128.0      BB_4   
  4      64      4744.0    2000.0      128.0      BB_4   
  5      64      4872.0    2000.0      128.0      BB_4   
  6      64      5000.0    2000.0      128.0      BB_4   
  7      64      5128.0    2000.0      128.0      BB_4   
  8      64      5256.0    2000.0      128.0      BB_4   
  9      64      5384.0    2000.0      128.0      BB_4   
  10    64      6488.0    2000.0      128.0      BB_8   
  11    64      6616.0    2000.0      128.0      BB_8   
  12    64      6744.0    2000.0      128.0      BB_8   
  13    64      6872.0    2000.0      128.0      BB_8   
  14    64      7000.0    2000.0      128.0      BB_8   
  15    64      7128.0    2000.0      128.0      BB_8   
  16    64      7256.0    2000.0      128.0      BB_8   
  17    64      7384.0    2000.0      128.0      BB_8   
Field information:
  FldID  Code  Name            RA            Dec            SrcID
  0      D      J0925+0019      09:25:07.82  +000.19.13.933  0   
  1      NONE  SN2010FZ        09:42:04.77  +000.19.51.000  1   
  2      K      3C286            13:31:08.29  +030.30.32.959  2   
Antennas (27):
  ID    Name  Station  Diam.(m)  Lat.          Long.
  0    ea01  W09      25.0      +000.00.00.0  +000.00.00.0
  1    ea02  E02      25.0      +033.53.51.0  -107.37.25.2
  2    ea03  E09      25.0      +033.54.01.1  -107.37.04.4
  3    ea04  W01      25.0      +033.53.53.6  -107.36.45.1
  4    ea05  W08      25.0      +033.54.00.5  -107.37.05.9
  5    ea06  N06      25.0      +033.53.53.0  -107.37.21.6
  6    ea08  N01      25.0      +033.54.10.3  -107.37.06.9
  7    ea09  E06      25.0      +033.54.01.8  -107.37.06.0
  8    ea10  N03      25.0      +033.53.57.7  -107.36.55.6
  9    ea11  E04      25.0      +033.54.04.8  -107.37.06.3
  10    ea12  E08      25.0      +033.53.59.7  -107.37.00.8
  11    ea13  N07      25.0      +033.53.55.1  -107.36.48.9
  12    ea14  E05      25.0      +033.54.12.9  -107.37.07.2
  13    ea15  W06      25.0      +033.53.58.8  -107.36.58.4
  14    ea16  W02      25.0      +033.53.56.4  -107.37.15.6
  15    ea17  W07      25.0      +033.54.00.9  -107.37.07.5
  16    ea18  N09      25.0      +033.53.54.8  -107.37.18.4
  17    ea19  W04      25.0      +033.54.19.0  -107.37.07.8
  18    ea20  N05      25.0      +033.53.59.1  -107.37.10.8
  19    ea21  E01      25.0      +033.54.08.0  -107.37.06.7
  20    ea22  N04      25.0      +033.53.59.2  -107.37.05.7
  21    ea23  E07      25.0      +033.54.06.1  -107.37.06.5
  22    ea24  W05      25.0      +033.53.56.5  -107.36.52.4
  23    ea25  N02      25.0      +033.53.57.8  -107.37.13.0
  24    ea26  W03      25.0      +033.54.03.5  -107.37.06.2
  25    ea27  E03      25.0      +033.54.00.1  -107.37.08.9
  26    ea28  N08      25.0      +033.54.00.5  -107.37.02.8
</pre>
The C-band data of interest is contained in scans 6-44 and spans spectral windows 2 to 17.
 
We use the {{importevla}} task to convert the SDM dataset from the archive to a CASA Measurement Set (MS).
 
<source lang="python">
# In CASA
importevla(asdm='/lustre/smyers/AS1015/AS1015_sb1658169_1.55388.89474846065', \
          vis='SN2010FZ_filled.ms',online=True,flagzero=True, \
          shadow=True,applyflags=False,tbuff=1.5,flagbackup=False)
</source>
Note that this will produce a warning in the logger and terminal:
<pre>
SDMDataObjectReaderException : Could not open file '/home/sandrock3/smyers/AS1015/AS1015_sb1658169_1.55388.89474846065/ASDMBinary/uid___evla_bdf_X1'. The message was 'No such file or directory'
WARNING ERROR: No data retrieving for this SDM mainTable row
</pre>
This indicates that there was a missing scan in the SDM binary tables. Later on we will find out whats missing!
 
Here we had the task create (but not apply) the online flagging commands, plus flags for zero-clipping and shadowing.  The timeranges for the online flags were extended by 1.5sec (the integration time was 1sec) to account for some timing mismatches present in the EVLA data at this time.  These online flags indicated times where the antennas were not on source (e.g. slewing) or had other detectable faults. The created flagging commands will be stored in the <tt>FLAG_CMD</tt> MS table and can be applied later.  Note that if you set <tt>applyflags=True</tt> here then after filling the task will go ahead and apply the flags for you.
 
For the purposes of this exercise, in order to save time and disk space, we have turned off the automatic creation of flag column backups by setting <tt>flagbackup=False</tt>.  If we make a mistake and need to recover flags then we will have to rerun all previous commands.  We recommend that for real data processing that you leave the default value <tt>flagbackup=True</tt> in this and subsequent tasks.
 
You now have a MS called <tt>SN2010FZ_filled.ms</tt> in your working area.  This should be 37GB like the SDM.
 
== Application of Online Flags and Averaging your MS ==
 
If you are starting from the filled MS, you can find this at the AOC at:
<pre>/lustre/smyers/AS1015/SN2010FZ_filled.ms</pre>
Again, use the actual location of this file for your system.
 
Summarize the filled MS:
<source lang="python">
# In CASA
listobs('SN2010FZ_filled.ms')
</source>
 
The scan part of the listing (with intent columns removed):
<pre>
  Date        Timerange (UTC)          Scan  FldId FieldName nVis  Int(s)  SpwIds     
  11-Jul-2010/21:28:28.5 - 21:29:26.5    1      0 J0925+0019  41418  1        [0, 1]                     
              21:29:27.5 - 21:30:56.5    2      0 J0925+0019  63180  1        [0, 1]                     
              21:30:57.5 - 21:32:26.5    3      0 J0925+0019  63180  1        [0, 1]                     
              21:32:27.5 - 21:33:56.5    4      0 J0925+0019  63180  1        [0, 1]                     
              21:33:57.5 - 21:34:56.5    5      0 J0925+0019  42120  1        [0, 1]                     
              21:34:57.5 - 21:35:55.5    6      0 J0925+0019  331344 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:35:56.5 - 21:38:25.5    7      0 J0925+0019  842400 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:38:26.5 - 21:39:55.5    9      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:39:56.5 - 21:41:24.5    10      1 SN2010FZ    499824 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:41:25.5 - 21:42:54.5    11      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:42:55.5 - 21:44:24.5    12      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:44:25.5 - 21:45:54.5    13      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:45:55.5 - 21:47:23.5    14      1 SN2010FZ    499824 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:47:24.5 - 21:47:53.5    15      1 SN2010FZ    168480 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:47:54.5 - 21:49:23.5    16      0 J0925+0019  505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:49:24.5 - 21:50:53.5    17      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:50:54.5 - 21:52:23.5    18      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:52:24.5 - 21:53:52.5    19      1 SN2010FZ    499824 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:53:53.5 - 21:55:22.5    20      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:55:23.5 - 21:56:52.5    21      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:56:53.5 - 21:58:22.5    22      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:58:23.5 - 21:58:52.5    23      1 SN2010FZ    168480 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              21:58:53.5 - 22:00:21.5    24      0 J0925+0019  499824 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:00:22.5 - 22:01:51.5    25      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:01:52.5 - 22:03:21.5    26      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:03:22.5 - 22:04:51.5    27      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:04:52.5 - 22:06:20.5    28      1 SN2010FZ     499824 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:06:21.5 - 22:07:50.5    29      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:07:51.5 - 22:09:20.5    30      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:09:21.5 - 22:09:50.5    31      1 SN2010FZ    168480 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:09:51.5 - 22:11:20.5    32      0 J0925+0019  505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:11:21.5 - 22:12:49.5    33      1 SN2010FZ    499824 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:12:50.5 - 22:14:19.5    34      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:14:20.5 - 22:15:49.5    35      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:15:50.5 - 22:17:19.5    36      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:17:20.5 - 22:18:48.5    37      1 SN2010FZ    499824 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:18:49.5 - 22:20:18.5    38      1 SN2010FZ    505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:20:19.5 - 22:20:48.5    39      1 SN2010FZ    168480 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:20:49.5 - 22:22:18.5    40      0 J0925+0019  505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:22:19.5 - 22:23:47.5    41      2 3C286        499824 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:23:48.5 - 22:25:17.5    42      2 3C286        505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
              22:25:18.5 - 22:26:47.5    43      2 3C286        505440 1        [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]
</pre>
Careful examination shows that scan 8 is missing but from the time ranges that the data has been merged into scan 7.  This sort of correlator back end data capture issues was occasionally seen during 2010.  Hopefully it will not affect our data but we should keep an eye out for problems with scan 7.  In any event this was likely the source of the error seen during the importevla step.
 
There also does not seem to be data for scan 44.
[[Image:plotSN2010FZ_flagcmd.png|200px|thumb|right|online flags plotted from flagcmd]]
You can list the commands stored in the <tt>FLAG_CMD</tt> table using {{flagcmd}}.
<source lang="python">
# In CASA
flagcmd(vis='SN2010FZ_filled.ms',flagmode='table',optype='list')
</source>
These will go to the logger.
 
You can plot the commands stored in the <tt>FLAG_CMD</tt> table using {{flagcmd}}.
<source lang="python">
# In CASA
myrows = range(285)
flagcmd(vis='SN2010FZ_filled.ms',flagmode='table',optype='plot',flagrows=myrows)
</source>
This will bring up a <tt>matplotlib</tt> plotter.  You can have it plot to a PNG file instead:
<source lang="python">
# In CASA
flagcmd(vis='SN2010FZ_filled.ms',flagmode='table',optype='plot',flagrows=myrows,outfile='plotSN2010FZ_flagcmd.png')
</source>
NOTE: plotting the flags will not work in Version 3.2.1 (you will get a blank plot) unless you restrict the rows to
include only the commands that have antenna and timerange. Later versions (3.3.0 onward) should work without row
selection.
 
The flags as plotted in the figure to the above right look normal. 
They are color-coded by REASON, and you see OFF_SOURCE flags between scans, and the occasional SUBREFLECTOR flag also between scans (most likely after band changes when the subreflector rotates to
pick up the new feed on the ring, some are slower than others).  What you watch for here are long blocks of unexpected flags like FOCUS, which might be false alarms and cause you to flag too much data.  In that case look at the data itself in {{plotms}} (see below for examples) before applying the flags.
 
To apply the flags also use {{flagcmd}}:
<source lang="python">
# In CASA
flagcmd(vis='SN2010FZ_filled.ms',flagmode='table',optype='apply',flagbackup=False)
</source>
This can take a while for our 37GB dataset.  It took 20min on my workstation.
 
With the known bad data flagged, we can now split out the data we want and also average down in time to make a smaller MS.
For D-configuration (max baselines 1km) we can safely average to 3s or even 10s to reduce dataset size:
<source lang="python">
# In CASA
split(vis='SN2010FZ_filled.ms',outputvis='SN2010FZ_filled10s.ms',datacolumn='data',timebin='10s')
</source>
This can also take a while for our 37GB dataset.  It took 20min on my workstation.
 
You now have a MS called <tt>SN2010FZ_filled10s.ms</tt> in your working area.  This should be 3.2GB in size.
 
== Examining and Flagging your Averaged MS ==
 
If you are starting from the pre-flagged averaged split MS, you can find this at the AOC at:
<pre>/lustre/smyers/AS1015/SN2010FZ_filled10s.ms</pre>
 
We use {{listobs}} to summarize our new MS:
<source lang="python">
# In CASA
listobs('SN2010FZ_filled10s.ms')
</source>
Scan 6 is a dummy scan so we will use scans 7 to 44 when we process our data.
 
[[Image:plotSN2010FZ_plotants.png|200px|thumb|right|plotants plotter]]
To plot up the antenna positions in the array:
<source lang="python">
# In CASA
plotants('SN2010FZ_filled10s.ms')
</source>
NOTE: if after this point or any other you get table locks, use {{clearstat}} to clear them:
<source lang="python">
# In CASA
clearstat
</source>
 
Now we examine the MS looking for bad data to flag. We will use {{plotms}} to bring up an interactive GUI that will display 2-D Y vs.X style line plots. NOTE: We do not recommend using the editing/flagging features of plotms.  It is very easy to mess up your data that way.  Also, to improve speed we will be restricting the scope of plotting so most box/flag operations would not get rid of all the bad data.  We will instead use plotms to identify bad data and then use flagcmd to flag it.  This will also allow full scripting of the flagging.
 
NOTE: If you need an introduction to {{plotms}}, see:
* [[Data flagging with plotms]]
* [[Averaging data in plotms]]
* [[What's the difference between Antenna1 and Antenna2? Axis definitions in plotms]]
 
WARNING: The '''Flag''' [[Image:FlagThoseData.png]] button on the plotms GUI is close to other buttons you will be using, in particular the one that gets rid of boxes you have drawn.  Be careful and don't hit the '''Flag''' button by mistake!
 
The useful spw are 2~17. To get an idea of the data layout, plot a single baseline/channel versus time:
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_filled10s.ms',field='',spw='2~17:31~31', \
      antenna='ea01&ea02',correlation='RR,LL',xaxis='time',yaxis='amp')
</source>
 
[[Image:plotSN2010FZ_plotms_ants.png|200px|thumb|right|plotms ant2 vs ea01]]
Look for bad antennas by picking the last field and plotting baselines versus antenna <tt>ea01</tt>:
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_filled10s.ms',field='2',spw='2~17:31~31', \
      antenna='ea01',correlation='RR,LL',xaxis='antenna2',yaxis='amp')
</source>
You should be able to see that antenna 11 (= ea13) is bad (very low amplitude, it has no C-band receiver!) and that some of the spectral windows on 15 and 23 (ea17,ea25) are also on the low side.  Boxing with the '''Mark Regions''' [[Image:MarkRegionsButton.png]] tool and using the '''Locate''' [[File:casaplotms-locate-tool.png]] tool will show in the logger that spw 10~17 are suspect for these antennas.
 
[[Image:plotSN2010FZ_plotms_ea02.png|200px|thumb|right|plotms ea02 vs frequency]]
Now look at the bandpass for ea02 - it is in the inner core and a prospective reference antenna. Exclude ea13 using negation in the selection:
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_filled10s.ms',field='2',spw='2~17', \
      antenna='ea02;!ea13',correlation='RR,LL',xaxis='frequency',yaxis='amp')
</source>
There is clearly less data for spw 11, and use of Locate shows spw 11 data only for ea02,ea03,04,08,09,11,12. We will later delete this incomplete spw.  Note also the very strong RFI spike at 6614MHz (spw 10 ch 63) with clear ringing contaminating both spw 10 and 11. There is also a tremendous roll-off in spw 10.  We will drop these spectral window when we process the data.
 
[[Image:plotSN2010FZ_plotms_ea02ea20.png|200px|thumb|right|plotms ea02&ea20 iteration phase vs frequency]]
We can also step through the baselines to our antenna using iteraxis - use the ">" button to step through:
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_filled10s.ms',field='2',spw='2~17',antenna='ea02;!ea13', \
      correlation='RR,LL',xaxis='frequency',yaxis='amp',iteraxis='baseline')
</source>
This will make it easier to isolate the bad antennas. Now plot the phases, iterating through baselines to ea02:
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_filled10s.ms',field='2',spw='2~17',antenna='ea02;!ea13', \
      correlation='RR,LL',xaxis='frequency',yaxis='phase',iteraxis='baseline')
</source>
You see the slopes due to residual delays. Mostly a turn or less over a 128MHz subband, but there are some outliers.
Step through to ea20.  You see that there is a very large delay in RR (via locate) for
the first baseband (spw 0~7).  We will delete this (will also delete LL so there are no orphan polarization products).
Note ea17 and ea25 baselines drop close to zero in the middle of upper baseband (e.g. plot 'ea17&ea25') so we will delete these.
 
To carry out flagging, we again use {{flagcmd}} in the mode where it takes a list of command strings:
<source lang="python">
# In CASA
flaglist = ['antenna="ea13"',
            'antenna="ea17" spw="10~17"',
            'antenna="ea25" spw="10~17"',
            'antenna="ea20" spw="2~9"']
flagcmd(vis='SN2010FZ_filled10s.ms',flagmode='cmd',command=flaglist,optype='apply',flagbackup=False)
</source>
These commands will be carried out as well as being added to the FLAG_CMD table (marked as applied).
 
Plot the data again now having flagged:
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_filled10s.ms',field='2',spw='2~17',antenna='ea02', \
      correlation='RR,LL',xaxis='frequency',yaxis='amp',scan='7~44')
</source>
 
[[Image:plotSN2010FZ_plotms_ea02fld0.png|200px|thumb|right|plotms field 0 ea02 amp vs frequency]]
Now our phase calibrator - it is weaker, and we now start to see the RFI:
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_filled10s.ms',field='0',spw='2~17',antenna='ea02', \
      correlation='RR,LL',xaxis='frequency',yaxis='amp',scan='7~44')
</source>
Use the Zoom feature, Mark rectangles and use Locate to identify the frequency/channel of RFI. In particular, we note in our analysis:
* 6614MHz (spw 10 ch 63) super strong
* 6772-6778MHz (spw 12 ch 14-17)
* 7260-7264MHz (spw 16 ch 2-4)
* 7314-7340MHz (spw 16 ch 29-42)
* 7402-7418MHz (spw 17 ch  9-17)
* 7458-7466MHz (spw 17 ch 37-41)
* 7488MHz (spw 17 ch 52)
 
If you plot all antennas and avoid the band edges you see spw 16,17 are pretty wiped out:
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_filled10s.ms',field='0',spw='2~17:4~59',antenna='', \
      correlation='RR,LL',xaxis='frequency',yaxis='amp',scan='7~44')
</source>
For now we will not flag these channels but note them to mask out when creating continuum calibration tables and images.
 
Finally, split off the good scans and spw, this will allow us to work on the data without having to start completely over if we mess something up badly as well as letting us do simpler selection:
<source lang="python">
# In CASA
split(vis='SN2010FZ_filled10s.ms',outputvis='SN2010FZ_flagged10s.ms',datacolumn='data',spw='2~9,12~17',scan='7~44')
</source>
 
You now have a MS called <tt>SN2010FZ_flagged10s.ms</tt> in your working area.  This should be 2.8GB in size.
 
== Calibration ==
 
Summarize the split flagged MS:
<source lang="python">
# In CASA
listobs('SN2010FZ_flagged10s.ms')
</source>
In the logger we see:
<pre>
================================================================================
          MeasurementSet Name:  /home/sandrock2/smyers/casa/tutorials/SN2010FZ/SN2010FZ_flagged10s.ms      MS Version 2
================================================================================
  Observer: Dr. Alicia M. Soderberg    Project: T.B.D. 
Observation: EVLA
Data records: 1374548      Total integration time = 3042 seconds
  Observed from  11-Jul-2010/21:36:01.0  to  11-Jul-2010/22:26:43.0 (UTC)
 
  ObservationID = 0        ArrayID = 0
  Date        Timerange (UTC)          Scan  FldId FieldName nVis  Int(s)  SpwIds      ScanIntent
  11-Jul-2010/21:36:01.0 - 21:38:20.5    7      0 J0925+0019  73710  9.93    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]CALIBRATE_PHASE#UNSPECIFIED
              21:38:44.0 - 21:39:51.0    9      1 SN2010FZ    39312  9.11    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              21:40:01.0 - 21:41:20.5    10      1 SN2010FZ    44226  9.89    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              21:41:30.0 - 21:42:50.0    11      1 SN2010FZ    44226  10      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              21:43:00.0 - 21:44:20.0    12      1 SN2010FZ    44226  10      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              21:44:30.0 - 21:45:50.0    13      1 SN2010FZ    44226  10      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              21:46:00.0 - 21:47:19.5    14      1 SN2010FZ    44226  9.89    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              21:47:29.0 - 21:47:48.5    15      1 SN2010FZ    14742  9.67    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              21:48:12.0 - 21:49:18.5    16      0 J0925+0019  39312  8.92    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]CALIBRATE_PHASE#UNSPECIFIED
              21:49:42.0 - 21:50:49.0    17      1 SN2010FZ    39312  9.11    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              21:50:59.0 - 21:52:19.0    18      1 SN2010FZ    44226  10      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              21:52:29.0 - 21:53:48.5    19      1 SN2010FZ    44226  9.89    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              21:53:58.0 - 21:55:18.0    20      1 SN2010FZ    44226  10      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              21:55:28.0 - 21:56:48.0    21      1 SN2010FZ    44226  10      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              21:56:58.0 - 21:58:18.0    22      1 SN2010FZ    44226  10      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              21:58:28.0 - 21:58:47.0    23      1 SN2010FZ    14742  9.33    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              21:59:11.0 - 22:00:17.0    24      0 J0925+0019  39312  8.93    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]CALIBRATE_PHASE#UNSPECIFIED
              22:00:40.0 - 22:01:47.0    25      1 SN2010FZ    39312  9.12    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              22:01:57.0 - 22:03:17.0    26      1 SN2010FZ    44226  10      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              22:03:27.0 - 22:04:47.0    27      1 SN2010FZ    44226  10      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              22:04:57.0 - 22:06:16.5    28      1 SN2010FZ    44226  9.89    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              22:06:26.0 - 22:07:46.0    29      1 SN2010FZ    44226  10      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              22:07:56.0 - 22:09:16.0    30      1 SN2010FZ    44226  10      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              22:09:26.0 - 22:09:45.0    31      1 SN2010FZ    14742  9.33    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              22:10:08.5 - 22:11:15.0    32      0 J0925+0019  39312  8.94    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]CALIBRATE_PHASE#UNSPECIFIED
              22:11:38.5 - 22:12:45.5    33      1 SN2010FZ    39312  9.11    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              22:12:55.0 - 22:14:15.0    34      1 SN2010FZ    44226  10      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              22:14:25.0 - 22:15:45.0    35      1 SN2010FZ    44226  10      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              22:15:55.0 - 22:17:15.0    36      1 SN2010FZ    44226  10      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              22:17:25.0 - 22:18:44.5    37      1 SN2010FZ    44226  9.89    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              22:18:54.0 - 22:20:14.0    38      1 SN2010FZ    44226  10      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              22:20:24.0 - 22:20:43.5    39      1 SN2010FZ    14742  9.6      [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
              22:21:06.5 - 22:22:13.5    40      0 J0925+0019  38584  8.96    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]CALIBRATE_PHASE#UNSPECIFIED
              22:25:13.0 - 22:25:13.0    42      2 3C286        770    2.65    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]CALIBRATE_BANDPASS#UNSPECIFIED,CALIBRATE_AMPLI#UNSPECIFIED
              22:25:23.0 - 22:26:43.0    43      2 3C286        42812  9.62    [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]CALIBRATE_BANDPASS#UNSPECIFIED,CALIBRATE_AMPLI#UNSPECIFIED
          (nVis = Total number of time/baseline visibilities per scan)
Fields: 3
  ID  Code Name        RA            Decl          Epoch  SrcId nVis 
  0    D    J0925+0019  09:25:07.8150 +00.19.13.9334 J2000  0    230230
  1    NONE SN2010FZ    09:42:04.7700 +00.19.51.0000 J2000  1    1100736
  2    K    3C286        13:31:08.2880 +30.30.32.9589 J2000  2    43582 
  (nVis = Total number of time/baseline visibilities per field)
Spectral Windows:  (14 unique spectral windows and 1 unique polarization setups)
  SpwID  #Chans Frame Ch1(MHz)    ChanWid(kHz)TotBW(kHz)  Ref(MHz)    Corrs         
  0          64 TOPO  4488        2000        128000      4488        RR  RL  LR  LL 
  1          64 TOPO  4616        2000        128000      4616        RR  RL  LR  LL 
  2          64 TOPO  4744        2000        128000      4744        RR  RL  LR  LL 
  3          64 TOPO  4872        2000        128000      4872        RR  RL  LR  LL 
  4          64 TOPO  5000        2000        128000      5000        RR  RL  LR  LL 
  5          64 TOPO  5128        2000        128000      5128        RR  RL  LR  LL 
  6          64 TOPO  5256        2000        128000      5256        RR  RL  LR  LL 
  7          64 TOPO  5384        2000        128000      5384        RR  RL  LR  LL 
  8          64 TOPO  6744        2000        128000      6744        RR  RL  LR  LL 
  9          64 TOPO  6872        2000        128000      6872        RR  RL  LR  LL 
  10        64 TOPO  7000        2000        128000      7000        RR  RL  LR  LL 
  11        64 TOPO  7128        2000        128000      7128        RR  RL  LR  LL 
  12        64 TOPO  7256        2000        128000      7256        RR  RL  LR  LL 
  13        64 TOPO  7384        2000        128000      7384        RR  RL  LR  LL 
Sources: 42
  ID  Name        SpwId RestFreq(MHz)  SysVel(km/s)
  0    J0925+0019  0    -              -           
  0    J0925+0019  1    -              -           
  0    J0925+0019  2    -              -           
  0    J0925+0019  3    -              -           
  0    J0925+0019  4    -              -           
  0    J0925+0019  5    -              -           
  0    J0925+0019  6    -              -           
  0    J0925+0019  7    -              -           
  0    J0925+0019  8    -              -           
  0    J0925+0019  9    -              -           
  0    J0925+0019  10    -              -           
  0    J0925+0019  11    -              -           
  0    J0925+0019  12    -              -           
  0    J0925+0019  13    -              -           
  1    SN2010FZ    0    -              -           
  1    SN2010FZ    1    -              -           
  1    SN2010FZ    2    -              -           
  1    SN2010FZ    3    -              -           
  1    SN2010FZ    4    -              -           
  1    SN2010FZ    5    -              -           
  1    SN2010FZ    6    -              -           
  1    SN2010FZ    7    -              -           
  1    SN2010FZ    8    -              -           
  1    SN2010FZ    9    -              -           
  1    SN2010FZ    10    -              -           
  1    SN2010FZ    11    -              -           
  1    SN2010FZ    12    -              -           
  1    SN2010FZ    13    -              -           
  2    3C286        0    -              -           
  2    3C286        1    -              -           
  2    3C286        2    -              -           
  2    3C286        3    -              -           
  2    3C286        4    -              -           
  2    3C286        5    -              -           
  2    3C286        6    -              -           
  2    3C286        7    -              -           
  2    3C286        8    -              -           
  2    3C286        9    -              -           
  2    3C286        10    -              -           
  2    3C286        11    -              -           
  2    3C286        12    -              -           
  2    3C286        13    -              -           
Antennas: 27:
  ID  Name  Station  Diam.    Long.        Lat.       
  0    ea01  W09      25.0 m  -107.37.25.2  +33.53.51.0 
  1    ea02  E02      25.0 m  -107.37.04.4  +33.54.01.1 
  2    ea03  E09      25.0 m  -107.36.45.1  +33.53.53.6 
  3    ea04  W01      25.0 m  -107.37.05.9  +33.54.00.5 
  4    ea05  W08      25.0 m  -107.37.21.6  +33.53.53.0 
  5    ea06  N06      25.0 m  -107.37.06.9  +33.54.10.3 
  6    ea08  N01      25.0 m  -107.37.06.0  +33.54.01.8 
  7    ea09  E06      25.0 m  -107.36.55.6  +33.53.57.7 
  8    ea10  N03      25.0 m  -107.37.06.3  +33.54.04.8 
  9    ea11  E04      25.0 m  -107.37.00.8  +33.53.59.7 
  10  ea12  E08      25.0 m  -107.36.48.9  +33.53.55.1 
  11  ea13  N07      25.0 m  -107.37.07.2  +33.54.12.9 
  12  ea14  E05      25.0 m  -107.36.58.4  +33.53.58.8 
  13  ea15  W06      25.0 m  -107.37.15.6  +33.53.56.4 
  14  ea16  W02      25.0 m  -107.37.07.5  +33.54.00.9 
  15  ea17  W07      25.0 m  -107.37.18.4  +33.53.54.8 
  16  ea18  N09      25.0 m  -107.37.07.8  +33.54.19.0 
  17  ea19  W04      25.0 m  -107.37.10.8  +33.53.59.1 
  18  ea20  N05      25.0 m  -107.37.06.7  +33.54.08.0 
  19  ea21  E01      25.0 m  -107.37.05.7  +33.53.59.2 
  20  ea22  N04      25.0 m  -107.37.06.5  +33.54.06.1 
  21  ea23  E07      25.0 m  -107.36.52.4  +33.53.56.5 
  22  ea24  W05      25.0 m  -107.37.13.0  +33.53.57.8 
  23  ea25  N02      25.0 m  -107.37.06.2  +33.54.03.5 
  24  ea26  W03      25.0 m  -107.37.08.9  +33.54.00.1 
  25  ea27  E03      25.0 m  -107.37.02.8  +33.54.00.5 
  26  ea28  N08      25.0 m  -107.37.07.5  +33.54.15.8 
</pre>
Note that the spws are re-numbered to 0-13.
 
Prepare the MS for calibration by adding scratch columns. This will take a few minutes.
<source lang="python">
# In CASA
clearcal('SN2010FZ_flagged10s.ms')
</source>
 
=== Setting the flux density scale ===
 
Before calibrating, we put a model for flux calibration source 3C286 into the MS (in the MODEL_DATA column we just created).
Have it set the flux on a per-channel basis.
NOTE: This uses the AOC Unix location for casapy models (yours may be different):
<source lang="python">
# In CASA
setjy(vis='SN2010FZ_flagged10s.ms',field='2',spw='',scalebychan=True, \
      modimage='/usr/lib64/casapy/data/nrao/VLA/CalModels/3C286_C.im')
</source>
It reports to logger that its about 7.7Jy at lower end to 5.7Jy at upper frequency limit.
 
=== Calibrating delays and bandpass ===
 
First, we do a phase-only calibration solution on a narrow range of channels in each spw on the bandpass/flux calibrator 3c286 to flatten them before solving for the bandpass. Note where we saw RFI in the higher spw and avoid those channels. The range 23~28 should work. Pick a refant near center - ea02 is a reasonable bet:
<source lang="python">
# In CASA
gaincal(vis='SN2010FZ_flagged10s.ms',caltable='calSN2010FZ.G0',field='2',spw='0~13:23~28', \
        gaintype='G',refant='ea02',calmode='p',solint='int',minsnr=3)
</source>
 
[[Image:plotSN2010FZ_plotcal_G0p1.png|200px|thumb|right|plotcal G0 phase ant 0~15]]
[[Image:plotSN2010FZ_plotcal_G0p2.png|200px|thumb|right|plotcal G0 phase ant 16~26]]
Plot the phase solutions (using full phase range instead of autorange)
<source lang="python">
# In CASA
plotcal(caltable='calSN2010FZ.G0',xaxis='time',yaxis='phase',iteration='antenna', \
        plotrange=[-1,-1,-180,180])
</source>
Step through the antenna-based solutions.  They look good (and fairly flat over the scans).
NOTE: If you want to make single-page multipanel plots (like those shown to the right), particularly for
hardcopy (where it only shows the first page), you can do:
<source lang="python">
# In CASA
plotcal(caltable='calSN2010FZ.G0',xaxis='time',yaxis='phase', \
        antenna='0~10,12~15',subplot=531,iteration='antenna', \
        plotrange=[-1,-1,-180,180],showgui=False,fontsize=6.0, \
        figfile='plotSN2010FZ_plotcal_G0p1.png')
plotcal(caltable='calSN2010FZ.G0',xaxis='time',yaxis='phase', \
        antenna='16~26',subplot=531,iteration='antenna', \
        plotrange=[-1,-1,-180,180],showgui=False,fontsize=6.0, \
        figfile='plotSN2010FZ_plotcal_G0p2.png')
</source>
 
We can now solve for the residual antenna-based delays that we saw in phase vs. frequency.
This uses the gaintype='K' option in gaincal (this is not documented in earlier releases, but is available in
Version 3.2.1 and later). Note that this currently does not do a "global fringe-fitting" solution for delays,
but instead does a baseline-based delay solution to all baselines to the refant, treating these
as antenna-based delays.  In most cases with high-enough S/N to get baseline-based delay solutions
this will suffice. We avoid the beginning of spw 0 due to the extreme roll-off (with loss of S/N) at the
starting edge.
<source lang="python">
# In CASA
gaincal(vis='SN2010FZ_flagged10s.ms',caltable='calSN2010FZ.K0',gaintable='calSN2010FZ.G0', \
        field='2',spw='0:8~59,1~13:4~59',gaintype='K', \
        refant='ea02',combine='scan',solint='inf',minsnr=3)
</source>
We pre-apply our initial phase table, and produce a new K-type caltable for input to bandpass calibration.
The delays found are sent to the terminal.  They are up to a few nanoseconds.
 
[[Image:plotSN2010FZ_plotcal_B0ea14.png|200px|thumb|right|plotcal B0 amp phase vs. freq for ea01]]
Now solve for the bandpass using the previous tables:
<source lang="python">
# In CASA
bandpass(vis='SN2010FZ_flagged10s.ms',caltable='calSN2010FZ.B0', \
        gaintable=['calSN2010FZ.G0','calSN2010FZ.K0'], \
        field='2',refant='ea02',solnorm=False, \
        bandtype='B', combine='scan', solint='inf', gaincurve=True)
</source>
'''WARNING''': You must set <tt>solnorm=False</tt> here or later on you will find some offsets
between spw due to how amplitude scaling adjusts weights internally during solving.
 
This is the first amplitude-scaling calibration that we do, and thus we have turned on the application
of gain-elevation curves (setting <tt>gaincurve=True</tt>).  If we were at higher frequency we would set
the <tt>opacity</tt> here also.  We will set these in every amplitude solve and application from now
on.
 
Now plot this, in amplitude then phase:
<source lang="python">
# In CASA
plotcal(caltable='calSN2010FZ.B0',xaxis='freq',yaxis='amp',iteration='antenna')
#
plotcal(caltable='calSN2010FZ.B0',xaxis='freq',yaxis='phase',iteration='antenna', \
        plotrange=[-1,-1,-180,180])
</source>
In the bandpass phase you no longer see the residual antenna delays (just residual spw phase offsets from
the delay solution registration) but there are some band edge effects.
Note that some antennas have a little strange bandpasses at upper end of lower baseband in spw 5,6,7
(e.g. ea14,ea16,ea17,ea25).
To plot amp and phase for a single antenna versus frequency (see plots at right):
<source lang="python">
# In CASA
plotcal(caltable='calSN2010FZ.B0',xaxis='freq',yaxis='amp', \
        antenna='ea14',subplot=211,figfile='')
plotcal(caltable='calSN2010FZ.B0',xaxis='freq',yaxis='phase', \
        antenna='ea14',subplot=212,plotrange=[-1,-1,-180,180], \
        figfile='plotSN2010FZ_plotcal_B0ea14.png')
</source>
 
Because our flux density calibrator 3C286 is bright enough, we were able to use this as the bandpass calibrator.
Since {{setjy}} put the correct spectrum for 3C286 into the MODEL_DATA column, our bandpass will reflect the
true bandpass of the instrument.  However, if for your observation you were unable to use a source of known spectrum
as the bandpass calibrator, then you would need to follow this bandpass with a second one on a source of known spectrum
in order to take out the spurious bandpass slope introduced by the (unknown) intrinsic spectral shape of your calibrator.
Running {{bandpass} with <tt>bandtype='BPOLY'</tt> and <tt>degamp=1</tt> should suffice to take out a slope albeit on a
per-antenna basis rather than over all the array.  This should work as long as you have enough S/N on your flux calibrator
to solve for two polynomial orders (might be hard if you are using very narrow bands at high frequency).
 
=== Final phase and amplitude calibration ===
 
[[Image:plotSN2010FZ_plotcal_G1p1.png|200px|thumb|right|plotcal G1 phase ant 0~15]]
[[Image:plotSN2010FZ_plotcal_G1p2.png|200px|thumb|right|plotcal G1 phase ant 16~26]]
Now calibrate phases for real with wider bandwidth. First the flux calibrator again, with a per-integration solution time:
<source lang="python">
# In CASA
gaincal(vis='SN2010FZ_flagged10s.ms',caltable='calSN2010FZ.G1', \
        gaintable=['calSN2010FZ.K0','calSN2010FZ.B0'], \
        field='2',refant='ea02',solnorm=F, spw='0~13:4~59', \
        solint='int',gaintype='G',calmode='p')
</source>
Next our phase calibrator, appending these solutions to previous table.
Exclude RFI channels here, and obtain one solution per scan:
<source lang="python">
# In CASA
gaincal(vis='SN2010FZ_flagged10s.ms',caltable='calSN2010FZ.G1',
        gaintable=['calSN2010FZ.K0','calSN2010FZ.B0'], \
        field='0',refant='ea02',solnorm=F, \
        spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
        solint='inf',gaintype='G',calmode='p',append=True)
</source>
The phases look reasonably connected:
<source lang="python">
# In CASA
plotcal(caltable='calSN2010FZ.G1',xaxis='time',yaxis='phase',iteration='antenna', \
        plotrange=[-1,-1,-180,180])
</source>
 
NOTE: If there were significant phase variations then you would also do short-timescale phase solutions on
the gain calibrator (field 0), as we did for 3C286, that you would apply only to that calibrator in order to get the amplitude
solutions correct.  You would still apply the per-scan phases to the target.  Because we have good phase stability
we will only do per-scan phase solutions on J0925+0019 and use that in both the amplitude solutions and to correct the target
phases.
 
Now solve for amplitudes on a per scan interval. Do these separately using gainfield so phases don't get
transferred across fields. Uses linear interpolation of the previously determined phases by default.
Pre-apply the gaincurve also:
<source lang="python">
# In CASA
gaincal(vis='SN2010FZ_flagged10s.ms', caltable='calSN2010FZ.G2', \
        gaintable=['calSN2010FZ.K0','calSN2010FZ.B0','calSN2010FZ.G1'], \
        gainfield=['2','2','2'], field='2',refant='ea02',solnorm=F,
        spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
        solint='inf',gaintype='G',calmode='a',gaincurve=True)
#
gaincal(vis='SN2010FZ_flagged10s.ms', caltable='calSN2010FZ.G2', \
        gaintable=['calSN2010FZ.K0','calSN2010FZ.B0','calSN2010FZ.G1'],\
        gainfield=['2','2','0'], field='0',refant='ea02',solnorm=F, \
        spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
        solint='inf',gaintype='G',calmode='a',gaincurve=True,append=True)
</source>
 
[[Image:plotSN2010FZ_plotcal_F2a1.png|200px|thumb|right|plotcal F2 amp ant 0~15]]
[[Image:plotSN2010FZ_plotcal_F2a2.png|200px|thumb|right|plotcal F2 amp ant 16~26]]
Since the flux on the gain calibrator is not scaled to its correct flux (but to 1.0Jy)
use fluxscale to transfer the amplitude gains from 3c286:
<source lang="python">
# In CASA
fluxscale(vis='SN2010FZ_flagged10s.ms',caltable='calSN2010FZ.G2', \
          fluxtable='calSN2010FZ.F2',reference='2',transfer='0')
</source>
The logger output gives:
<pre>
Found reference field(s): 3C286
Found transfer field(s):  J0925+0019
Flux density for J0925+0019 in SpW=0 is: 0.976065 +/- 0.000880351 (SNR = 1108.72, nAnt= 25)
Flux density for J0925+0019 in SpW=1 is: 0.97776 +/- 0.000811257 (SNR = 1205.24, nAnt= 25)
Flux density for J0925+0019 in SpW=2 is: 0.979643 +/- 0.00069687 (SNR = 1405.78, nAnt= 25)
Flux density for J0925+0019 in SpW=3 is: 0.980233 +/- 0.000838793 (SNR = 1168.62, nAnt= 25)
Flux density for J0925+0019 in SpW=4 is: 0.981888 +/- 0.000689379 (SNR = 1424.31, nAnt= 25)
Flux density for J0925+0019 in SpW=5 is: 0.982243 +/- 0.000813425 (SNR = 1207.54, nAnt= 21)
Flux density for J0925+0019 in SpW=6 is: 0.981002 +/- 0.000894034 (SNR = 1097.27, nAnt= 21)
Flux density for J0925+0019 in SpW=7 is: 0.98006 +/- 0.000672249 (SNR = 1457.88, nAnt= 21)
Flux density for J0925+0019 in SpW=8 is: 0.960763 +/- 0.000715631 (SNR = 1342.54, nAnt= 24)
Flux density for J0925+0019 in SpW=9 is: 0.957227 +/- 0.000781501 (SNR = 1224.86, nAnt= 24)
Flux density for J0925+0019 in SpW=10 is: 0.955039 +/- 0.000774128 (SNR = 1233.7, nAnt= 24)
Flux density for J0925+0019 in SpW=11 is: 0.953599 +/- 0.000944027 (SNR = 1010.14, nAnt= 24)
Flux density for J0925+0019 in SpW=12 is: 0.950831 +/- 0.000933345 (SNR = 1018.73, nAnt= 24)
Flux density for J0925+0019 in SpW=13 is: 0.946425 +/- 0.00101842 (SNR = 929.306, nAnt= 24)
</pre>
You may see slightly different numbers on your machine.
 
As it so happens, the derived flux for J0925+0019 is about 1 Jy (you can plot up the raw amplitudes for fields 0,2 and convince yourself this is indeed true and not a bug). The spectrum rises a bit to peak in spw 5 then falls again.
 
Plot these solutions:
<source lang="python">
# In CASA
plotcal(caltable='calSN2010FZ.F2',xaxis='time',yaxis='amp',iteration='antenna')
</source>
The gains on 3C286 are about 1 (the bandpass solution on 3C286 has absorbed the calibration from counts to Jy) but
fluxscale has adjusted the per-spw scale on J0925+0019 to get its correct spectrum rather than the assumed 1 Jy
flat spectrum.
 
== Applying the Calibration and Final Editing ==
 
Next we actually apply all our accumulated calibration tables. We apply these to the
calibration fields individually using the appropriate gainfields and interpolation for each:
* For 3C286 (field 2) we did short-timescale phase solutions and a single scan amplitude, so use "linear" and "nearest" interpolation respectively.
* For the nearby gain calibrator (field 0) we did only scan-based phase and amplitude solutions so we use "nearest" interpolation
* For the target source we use field 0 to calibrate field 1, so use "linear" interpolation. This takes a few minutes.
 
[[Image:plotSN2010FZ_plotms_applied_fld2.png|200px|thumb|right|plotms of 3C286 with calibration applied]]
<source lang="python">
# In CASA
applycal(vis='SN2010FZ_flagged10s.ms',field='2', \
        gaintable=['calSN2010FZ.K0','calSN2010FZ.B0','calSN2010FZ.G1','calSN2010FZ.F2'], \
        gainfield=['','','2','2'],interp=['nearest','nearest','linear','nearest'], \
        parang=False,calwt=F,gaincurve=T)
#
applycal(vis='SN2010FZ_flagged10s.ms',field='0', \
        gaintable=['calSN2010FZ.K0','calSN2010FZ.B0','calSN2010FZ.G1','calSN2010FZ.F2'], \
        gainfield=['','','0','0'], interp=['nearest','nearest','nearest','nearest'], \
        parang=False,calwt=F,gaincurve=T)
#
applycal(vis='SN2010FZ_flagged10s.ms',field='1', \
        gaintable=['calSN2010FZ.K0','calSN2010FZ.B0','calSN2010FZ.G1','calSN2010FZ.F2'], \
        gainfield=['','','0','0'], interp=['nearest','nearest','linear','linear'], \
        parang=False,calwt=F,gaincurve=T)
</source>
 
We can examine the corrected data on 3c286 using our RFI mask from above and avoiding band edges
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_flagged10s.ms',field='2', \
      spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
      correlation='RR,LL',xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
</source>
See figure above right.  There is clearly discrepant data visible spw 5 and 6, in particular for baseline ea17&ea25 (use the
'''Mark Regions''' [[Image:MarkRegionsButton.png]] tool on some of it and then use the '''Locate''' [[File:casaplotms-locate-tool.png]] tool), which gives a really strange response.  You can plot just this baseline to be sure:
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_flagged10s.ms',field='2', \
      spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
      antenna='ea17&ea25', \
      correlation='RR,LL',xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
</source>
You can exclude this through antenna negation:
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_flagged10s.ms',field='2', \
      spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
      antenna='!ea17&ea25', \
      correlation='RR,LL',xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
</source>
Then use '''Locate''' [[File:casaplotms-locate-tool.png]] for the other bad points, which seem to indicate spw 5,6,7 for ea14,ea16,ea17,ea25.
Exclude these and replot:
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_flagged10s.ms',field='2', \
      spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
      antenna='!ea14;!ea16;!ea17;!ea25', \
      correlation='RR,LL',xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
</source>
[[Image:plotSN2010FZ_plotms_appliedflags_fld2.png|200px|thumb|right|plotms cal applied flagged fld2]]
This now looks clean except for the RFI in the upper subbands.
 
Do flagging based on these:
<source lang="python">
# In CASA
flaglist = ['antenna="ea14,ea16,ea17,ea25" spw="5~7"']
flagcmd(vis='SN2010FZ_flagged10s.ms',flagmode='cmd',command=flaglist, \
        optype='apply',flagbackup=False)
</source>
 
Now replot the corrected data (you may have to force reload if you plotted same thing right before this):
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_flagged10s.ms',field='2', \
      spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
      correlation='RR,LL',xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
</source>
Looks good. If we were more diligent we would go back and recalibrate, but this looks good enough for now.
 
Plot the phase:
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_flagged10s.ms',field='2', \
      spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
      correlation='RR,LL',xaxis='frequency',yaxis='phase',ydatacolumn='corrected')
</source>
[[Image:plotSN2010FZ_plotms_appliedflags_fld2_phase.png|200px|thumb|right|plotms cal applied flagged fld2 phase]]
Note the characteristic "bowtie" pattern of the phases about the sub-band centers.
Here we can see the effect of the EVLA "delay clunking", where the delay steps through discrete values such that
the phase goes from -11deg to +11deg across the sub-band as the delay changes due to geometry.  This is D-configuration so the delays change slowly, it will change faster in wider configurations. As of Q3 2011 we have not enabled the corrections for this
in the EVLA system so you will always have this remaining delay error in your data.  In principle you could solve for delays on short timescales and take this out, in practice this in not possible for your weaker target source in any event.
 
Now lets plot the corrected data amplitude for the phase calibrator (field 0):
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_flagged10s.ms',field='0', \
      spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
      correlation='RR,LL',xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
</source>
[[Image:plotSN2010FZ_plotms_appliedflags_fld0_amp.png|200px|thumb|right|plotms cal applied flagged fld0 amp]]
[[Image:plotSN2010FZ_plotms_appliedflags_fld0_phase.png|200px|thumb|right|plotms cal applied flagged fld0 phase]]
You can see the roll-off pumping up noise at the baseband edges (about 8-16 channels worth). Also, we can see some RFI we missed:
* <6804MHz spw  8 below ch 30 lots of bad stuff (alot from ea18,ea22 but others too)
*  7168MHz spw 11 ch 20
*  pretty much all of spw 12,13
The ch 20 ones are all harmonics of the notorious 128MHz tone. NOTE: You can get the frequency of a RFI feature by looking at
the logger report from using the '''Locate''' [[File:casaplotms-locate-tool.png]] tool.
 
We will not flag these, but exclude them in imaging (so that more advanced students can try flagging these in detail or using
auto-flagging). A good channel selection string for imaging might be:
<pre>
spw = '0:16~59,1~6:4~59,7:4~54,8:30~59,9~10:4~59,11:4~19;21~59'
</pre>
Without further flagging, it may be best to drop spw 12-13 for imaging (we will do so from now on).
 
Plot again (including this selection for spw 0-11):
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_flagged10s.ms',field='0', \
      spw='0:16~59,1~6:4~59,7:4~54,8:30~59,9~10:4~59,11:4~19;21~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
      correlation='RR,LL',xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
</source>
Looks better.
[[Image:plotSN2010FZ_plotms_appliedflags_fld0_ampavg.png|200px|thumb|right|plotms cal applied flagged fld0 amp averaged]]
Now plot amplitudes for the corrected data averaged over baseline to see the source spectrum
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_flagged10s.ms',field='0', \
      spw='0:16~59,1~6:4~59,7:4~54,8:30~59,9~10:4~59,11:4~19;21~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
      correlation='RR,LL',avgbaseline=True,avgtime='60000s',
      xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
</source>
The last two sub-bands spw 12-13 give reasonable values, with only a tiny offset from spw 8-11.
There are also strange amplitude excursions, particularly in the low end of the first baseband.  These must be coming
from one or more scans.  You can iterate over scan to see the strange amplitudes (mostly in first scan):
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_flagged10s.ms',field='0', \
      spw='0:16~59,1~6:4~59,7:4~54,8:30~59,9~10:4~59,11:4~19;21~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
      correlation='RR,LL',avgbaseline=True,avgtime='600s',iteraxis='scan',
      xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
</source>
If we were interested in imaging the calibrator we would try to flag or fix this, but it shouldn't affect the
calibration transfer to our target very much (we did not see anomalous solutions in that first scan). 
 
We can also plot the corrected phase - looks fairly good
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_flagged10s.ms',field='0', \
      spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
      correlation='RR,LL',xaxis='frequency',yaxis='phase',ydatacolumn='corrected')
</source>
 
We can average over baseline and each scan:
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_flagged10s.ms',field='0', \
      spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
      correlation='RR,LL',avgbaseline=True,avgtime='600s',
      xaxis='frequency',yaxis='phase',ydatacolumn='corrected')
</source>
[[Image:plotSN2010FZ_plotms_appliedflags_fld0_phaseavg.png|200px|thumb|right|plotms cal applied flagged fld0 phase averaged]]
In this case we can see the residual effect of the EVLA "delay clunking" described above. It is reduced due to the averaging
that we applied, but it is still there.
 
[[Image:plotSN2010FZ_plotms_appliedflags_fld1.png|200px|thumb|right|plotms cal applied flagged fld1 amp averaged]]
You can look at the target source field='1' but there are lots of data so you will need to do alot of averaging.
For example:
<source lang="python">
# In CASA
plotms(vis='SN2010FZ_flagged10s.ms',field='1',avgtime='300s', \
      spw='0:16~59,1~6:4~59,7:4~54,8:30~59,9~10:4~59,11:4~19;21~59', \
      correlation='RR,LL',xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
</source>
Alas, the upper baseband still has lots of low level RFI.
 
Now split off the data for calibrators and target, to avoid later issues that can corrupt the MSes.
These will take a bit of time...field 1 took 18min on my workstation.
<source lang="python">
# In CASA
split(vis='SN2010FZ_flagged10s.ms',outputvis='SN2010FZ_split10s.ms',datacolumn='corrected',field='1')
#
split(vis='SN2010FZ_flagged10s.ms',outputvis='SN2010FZ_3c28610s.ms',datacolumn='corrected',field='2')
#
split(vis='SN2010FZ_flagged10s.ms',outputvis='SN2010FZ_J092510s.ms',datacolumn='corrected',field='0')
</source>
 
== Imaging ==
 
This is D-config data at C-band, look at the Obs Status Summary:
[http://evlaguides.nrao.edu/index.php?title=Observational_Status_Summary_-_Current]
Synthesized beam should be 12" at 6GHz with primary beam FOV 7.5arcmin (450").
Our data spans 4.5-7.5GHz : beam 9.6 at 7.5GHz and FOV 10' at 4.5GHz
A cellsize of 3" should work, with an imsize > 200 to cover 2xFWHM FOV
An imsize of 400+ will put the main beam inside the inner quarter safely
The Briggs robust (0.5) weighting is somewhere between uniform and natural
and will give reasonable resolution but still see some larger scale structure.
 
Due to the numerology of FFTW (which clean uses underneath for FFTs) optimal sizes,
imsize should be composite number with two and only two prime factors chosen from
2,3,5.  Taking into account the x1.2 padding that clean uses internally to the imsize
you give it (and 1.2=2*3/5), we choose 640 or 1280 as our imsize (640=2^7*5). Other
reasonable sets would be 405,1215,etc. (405=3^4*5) or 432,648,1296 (these are 2^n*3^m*5).
In practice, if you give it non-optimal values for imsize, you may find that the transforms
take a bit longer, which is noticeable if you are doing interactive clean.
 
WARNING: By default, a single-field nterms=1 clean does NOT use Cotton-Schwab (CS) clean to break
into major cycles going back to data residuals, it just does cleaning in a bunch of minor
cycles in the image plane.  This can give much poorer imaging quality in cases with poor
psf (snapshots) or in the case of complex emission structure (like ours) - clean tends to
diverge in this case.  You should explicitly set <tt>imagermode='csclean'</tt> in your
call to clean.  Also, in our case the psf is very good using mfs, so by default it will not
take many major cycle breaks.  We use the <tt>cyclefactor</tt> parameter to control this, which
sets the break threshold to be cyclefactor times the max psf sidelobe level (outside the main
peak).  We start at <tt>cyclefactor=1.5</tt> in a single spw, and ratchet it up to 4.5 when we
clean all the spw.  This seems to work ok.  Rule of thumb is if it is gobbling up many hundreds of
clean iterations in the minor cycles early on, increase cyclefactor.  Conversely, if your psf is poor
but you source structure is simple, you can reduce cyclefactor (e.g. below 1) to stop it from taking
lots of extra major cycles.
 
For more information on using {{clean}}, in particular on using the interactive GUI, see
[[EVLA_Continuum_Tutorial_3C391#Imaging]].
 
NOTE: If you are pressed for time, then you might want to jump ahead to
[[EVLA_6cmWideband_Tutorial_SN2010FZ#Cleaning_the_lower_baseband_using_two_MFS_Taylor_terms]] and will it is
cleaning read the other Imaging descriptions.
 
=== Cleaning a single spectral window ===
 
Let us start by interactively cleaning one of the lower baseband spw (spw 5 in this example).
NOTE: this first time will take a few minutes at start to create scratch columns
in the MS in case we want to do self-calibration later.
 
[[Image:viewSN2010FZ_spw5_clean640.png|200px|thumb|right|clean spw5 640x640]]
[[Image:viewSN2010FZ_spw5_clean1280.png|200px|thumb|right|clean spw5 1280x1280 before clean]]
[[Image:viewSN2010FZ_spw5_clean1280final.png|200px|thumb|right|clean spw5 1280x1280 restored image]]
 
<source lang="python">
# In CASA
clean(vis='SN2010FZ_split10s.ms',spw='5:4~59', \
      imagename='imgSN2010FZ10s_spw5_clean640', \
      mode='mfs',nterms=1,niter=10000,gain=0.1,threshold='0.0mJy', \
      psfmode='clark',imsize=[640,640],cell=['3.0arcsec'],stokes='I', \
      imagermode='csclean', cyclefactor=1.5, \
      weighting='briggs',robust=0.5,calready=True,interactive=True)
</source>
Start carefully by boxing the bright source and setting iterations to 10 at first
Gradually add more boxes and increase the number of iterations
Since this is not much more than a snapshot you see the six-fold sidelobe pattern
of the extended emission in the center of the map. This decreases as you clean
out this emission.
Stop cleaning when the residuals look like noise (and you cannot clearly see sources).
To stop click the red X button.
 
The top figure to the right shows a zoom in on the end state of the clean, where
we have marked a number of boxes and cleaned them out.
 
Note that there are some strange sidelobe patterns in lower left, possibly
from a source outside the image area.  We can make a bigger image starting from
our current model:
<source lang="python">
# In CASA
clean(vis='SN2010FZ_split10s.ms',spw='5:4~59', \
      imagename='imgSN2010FZ10s_spw5_clean1280', \
      mode='mfs',nterms=1,niter=10000,gain=0.1,threshold='0.0mJy', \
      psfmode='clark',imsize=[1280,1280],cell=['3.0arcsec'],stokes='I', \
      imagermode='csclean', cyclefactor=1.5, \
      modelimage='imgSN2010FZ10s_spw5_clean640.model', \
      weighting='briggs',robust=0.5,calready=True,interactive=True)
</source>
Sure enough, there is a bright source near the lower left (see middle panel at right).
Box it, clean it a bit, and look again.  There is a second source in the mid-left (track
it down by its sidelobes). Box this one, clean it a bit, and when satisfied stop.
 
You can use the CASA {{viewer}} to display the images that {{clean}} creates.  If you need more guidance
on using the viewer, see the [[http://casa.nrao.edu/CasaViewerDemo/casaViewerDemo.html CASAViewerDemo]] video. For now just bring up your restored image
directly:
<source lang="python">
# In CASA
viewer('imgSN2010FZ10s_spw5_clean1280.image')
</source>
The restored image is shown in bottom panel to the right.  I have chosen the Grayscale1 instead of default color
map as I prefer grayscale to false color for assessing image quality.
Check the rms of the residuals using the {{imstat}} task:
<source lang="python">
# In CASA
mystat = imstat('imgSN2010FZ10s_spw5_clean1280.residual')
print 'Residual standard deviation = '+str(mystat['sigma'][0])
</source>
I got 31.8uJy for mine.
 
=== Cleaning the lower baseband ===
 
[[Image:viewSN2010FZ_spw0to7_clean1280final.png|200px|thumb|right|clean spw0-7 restored image center]]
Now, image the entire lower baseband (spw 0-7).
Follow same iterative procedure as before, and get the best
residuals you can.  Because of the bandwidth and frequency
synthesis, the sidelobe pattern is different than before and
it is much easier to see fainter emission.
Be careful cleaning sources that lie near or on sidelobe splotches.
Clean the central emission region way down first to reduce the sidelobe
level before adding components in the sidelobe areas.
<source lang="python">
# In CASA
clean(vis='SN2010FZ_split10s.ms',spw='0:16~59,1~6:4~59,7:4~54', \
      imagename='imgSN2010FZ10s_spw0to7_clean1280', \
      mode='mfs',nterms=1,niter=10000,gain=0.1,threshold='0.0mJy', \
      psfmode='clark',imsize=[1280,1280],cell=['3.0arcsec'],stokes='I', \
      imagermode='csclean', cyclefactor=1.5, \
      weighting='briggs',robust=0.5,calready=True,interactive=True)
#
mystat = imstat('imgSN2010FZ10s_spw0to7_clean1280.residual')
print 'Residual standard deviation = '+str(mystat['sigma'][0])
</source>
I got 11.3uJy (and there is clearly structure left in the residual). To the right is a zoom-in on the center
of the restored image.
 
==== Cleaning the lower baseband using two MFS Taylor terms ====
 
The mfs nterms=2 option creates two "Taylor Term" images - an average intensity image (with suffix <tt>.image.tt0</tt>)
and a spectral slope image (with suffix <tt>.image.tt1</tt>) which is intensity x alpha (where alpha is spectral index).
For convenience there is a spectral index image (with suffix <tt>.image.alpha</tt>).  These Taylor expansions are with respect to the "Reference Frequency" of the image (by default the center frequency of the spw selected, but can be specified using the
reffreq parameter in clean). The convention for spectral index alpha is that
 
<math>
S \propto \nu^\alpha
</math>
 
so negative spectral indexes indicate a "steep" spectrum (falling with frequency).
 
[[Image:viewSN2010FZ_spw0to7_mfs2clean.png|200px|thumb|right|clean spw0-7 mfs nterms=2 in progress]]
[[Image:viewSN2010FZ_spw0to7_mfs2loadalpha.png|200px|thumb|right|clean spw0-7 mfs nterms=2 load alpha with LEL]]
[[Image:viewSN2010FZ_spw0to7_mfs2panelalpha.png|200px|thumb|right|clean spw0-7 mfs nterms=2 tt0 and alpha]]
Lets try using multi-frequency synthesis with nterms=2 on the lower baseband.
The dirty beam will have lower sidelobes so we turn up cyclefactor for csclean a bit:
<source lang="python">
# In CASA
clean(vis='SN2010FZ_split10s.ms',spw='0:16~59,1~6:4~59,7:4~54', \
      imagename='imgSN2010FZ10s_spw0to7_mfs2_clean1280', \
      mode='mfs',nterms=2,niter=10000,gain=0.1,threshold='0.0mJy', \
      psfmode='clark',imsize=[1280,1280],cell=['3.0arcsec'],stokes='I', \
      imagermode='csclean', cyclefactor=4.5, \
      weighting='briggs',robust=0.5,calready=True,interactive=True)
#
mystat = imstat('imgSN2010FZ10s_spw0to7_mfs2_clean1280.residual.tt0')
print 'Residual standard deviation = '+str(mystat['sigma'][0])
</source>
I got 10.5uJy (somewhat better looking than the nterms=1).
The top screenshot to the right shows an intermediate but early stage of cleaning where we are looking at
the central emission and cleaning it out slowly.
 
You can use the {{viewer}} to load the average intensity
<source lang="python">
# In CASA
viewer('imgSN2010FZ10s_spw0to7_mfs2_clean1280.image.tt0')
</source>
and then use the Open Data panel to load the spectral index
<tt>imgSN2010FZ10s_spw0to7_mfs2_clean1280.image.alpha</tt>
which can then be blinked (optionally plotted side-by-side using the Panel Display Options
panel to set 2 panels in the x direction). 
 
Note there is alot of noise in alpha in the low-intensity
regions, and thus filtering the alpha image based on the values in the tt0 image is desirable.
You can use the {{immath}} task to make this filtered alpha image explicitly, using a
Lattice Expression Language (LEL) expression:
<source lang="python">
# In CASA
immath(imagename=['imgSN2010FZ10s_spw0to7_mfs2_clean1280.image.alpha',
                  'imgSN2010FZ10s_spw0to7_mfs2_clean1280.image.tt0'],
      mode='evalexpr',
      expr='IM0[IM1>5.0E-5]',
      outfile='imgSN2010FZ10s_spw0to7_mfs2_clean1280.image.alpha.filtered')
</source>
This will use 50uJy (or 5 x the sigma we found) as the cutoff.
You can then view or manipulate the filtered alpha image as normal.
 
We can also use LEL to filter the alpha image on the intensity on-the-fly when we load this raster in the Open Data panel
by specifying a LEL string in the LEL box instead of selecting the image from the directory
listing.  The LEL string:
<pre>
'imgSN2010FZ10s_spw0to7_mfs2_clean1280.image.alpha'['imgSN2010FZ10s_spw0to7_mfs2_clean1280.image.tt0'>5.0E-05]
</pre>
will replicate what we did above. The middle figure to the right shows the Open Data panel
with our LEL string in it.  Just click the Raster button to load this.
 
The lower panel to the right shows the intensity and LEL-filtered alpha images side-by-side in the viewer, zoomed
in on the galaxy emission.  Mousing over the alpha shows spectral indexes ranging from -1 to +1 in the center, with
the brightest emission with alpha -0.7 in the knots in the disk.
 
=== Cleaning using both basebands combined ===
 
For the ultimate image, use the "clean" part of the upper baseband in addition
to the lower (use spw 0-11). We will use mfs with nterms=2 (if you try nterms=1
on this wide bandwidth you will get much poorer residuals). Because of the added
work and extra data involved, this will take much longer than our other runs of
clean.  Therefore, we will get a head start by doing a non-interactive clean using
the mask left from the previous clean (spw 0-7). We will insert a clean threshold
to limit runaway cleaning too far beneath the noise level. This took 40min on my
workstation.
<source lang="python">
# In CASA
clean(vis='SN2010FZ_split10s.ms', \
      spw='0:16~59,1~6:4~59,7:4~54,8:30~59,9~10:4~59,11:4~19;21~59', \
      imagename='imgSN2010FZ10s_spw0to11_mfs2_clean1280', \
      mode='mfs',nterms=2,niter=3000,gain=0.1,threshold='0.002mJy', \
      psfmode='clark',imsize=[1280,1280],cell=['3.0arcsec'],stokes='I', \
      imagermode='csclean', cyclefactor=4.5, \
      mask=['imgSN2010FZ10s_spw0to7_mfs2_clean1280.mask'], \
      weighting='briggs',robust=0.5,calready=True,interactive=False)
#
mystat = imstat('imgSN2010FZ10s_spw0to11_mfs2_clean1280.residual.tt0')
print 'Residual standard deviation = '+str(mystat['sigma'][0])
</source>
I got 8.6uJy for this effort (noticeably better than the lower baseband only results).
 
[[Image:viewSN2010FZ spw0to11_mfs2resid.png|200px|thumb|right|final residual and mask]]
Let us see if there is more to clean.  Bring this up in interactive mode:
<source lang="python">
# In CASA
clean(vis='SN2010FZ_split10s.ms', \
      spw='0:16~59,1~6:4~59,7:4~54,8:30~59,9~10:4~59,11:4~19;21~59', \
      imagename='imgSN2010FZ10s_spw0to11_mfs2_clean1280', \
      mode='mfs',nterms=2,niter=3000,gain=0.1,threshold='0.001mJy', \
      psfmode='clark',imsize=[1280,1280],cell=['3.0arcsec'],stokes='I', \
      imagermode='csclean', cyclefactor=4.5, \
      weighting='briggs',robust=0.5,calready=True,interactive=True)
</source>
I found a few more sources revealed in the outer parts of the image, and also more
emission around the galaxy disk in the center.  So I drew new boxes, extended the box
in the center, and did about 1000 new iterations.  At the end, what was left was dominated
by the error patterns from mis-calibration.  Only self-calibration will get rid of
these. Stop cleaning for now. See the figure to the right for the interactive display panel
showing final residuals and mask (changing the colormap to <tt>Greyscale 1</tt>).
 
Check the residual levels:
<source lang="python">
# In CASA
mystat = imstat('imgSN2010FZ10s_spw0to11_mfs2_clean1280.residual.tt0')
sigma = mystat['sigma'][0]
print 'Residual standard deviation = '+str(mystat['sigma'][0])
</source>
Now I get 8.3uJy.
 
== Analyzing the image and comparing with the Optical/Infrared ==
 
Lets see how close we got to expected noise and dynamic range.
<source lang="python">
# In CASA
mystat = imstat('imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0')
peak = mystat['max'][0]
print 'Image max flux = '+str(mystat['max'][0])
#
mystat = imstat('imgSN2010FZ10s_spw0to11_mfs2_clean1280.model.tt0')
total = mystat['sum'][0]
print 'Model total flux = '+str(mystat['sum'][0])
#
snr = peak/sigma
print 'SN2010FZ peak S/N = '+str(snr)
#
snr = total/sigma
print 'SN2010FZ total S/N = '+str(snr)
</source>
The output gives:
<pre>
Residual standard deviation = 8.33040657806e-06
Image max flux = 0.00995613634586
Model total flux = 0.0329372803525
SN2010FZ peak S/N = 1195.15611304
SN2010FZ total S/N = 3953.86228077
</pre>
 
What do we expect? If we do {{listobs}} on this MS we see the scans:
<pre>
Date        Timerange (UTC)          Scan  FldId FieldName nVis  Int(s) 
11-Jul-2010/21:38:44.0 - 21:39:51.0    9      0 SN2010FZ    39312  9.11 
            21:40:01.0 - 21:41:20.5    10      0 SN2010FZ    44226  9.89 
            21:41:30.0 - 21:42:50.0    11      0 SN2010FZ    44226  10   
            21:43:00.0 - 21:44:20.0    12      0 SN2010FZ    44226  10   
            21:44:30.0 - 21:45:50.0    13      0 SN2010FZ    44226  10   
            21:46:00.0 - 21:47:19.5    14      0 SN2010FZ    44226  9.89 
            21:47:29.0 - 21:47:48.5    15      0 SN2010FZ    14742  9.67 
            21:49:42.0 - 21:50:49.0    17      0 SN2010FZ    39312  9.11 
            21:50:59.0 - 21:52:19.0    18      0 SN2010FZ    44226  10   
            21:52:29.0 - 21:53:48.5    19      0 SN2010FZ    44226  9.89 
            21:53:58.0 - 21:55:18.0    20      0 SN2010FZ    44226  10   
            21:55:28.0 - 21:56:48.0    21      0 SN2010FZ    44226  10   
            21:56:58.0 - 21:58:18.0    22      0 SN2010FZ    44226  10   
            21:58:28.0 - 21:58:47.0    23      0 SN2010FZ    14742  9.33 
            22:00:40.0 - 22:01:47.0    25      0 SN2010FZ    39312  9.12 
            22:01:57.0 - 22:03:17.0    26      0 SN2010FZ    44226  10   
            22:03:27.0 - 22:04:47.0    27      0 SN2010FZ    44226  10   
            22:04:57.0 - 22:06:16.5    28      0 SN2010FZ    44226  9.89 
            22:06:26.0 - 22:07:46.0    29      0 SN2010FZ    44226  10   
            22:07:56.0 - 22:09:16.0    30      0 SN2010FZ    44226  10   
            22:09:26.0 - 22:09:45.0    31      0 SN2010FZ    14742  9.33 
            22:11:38.5 - 22:12:45.5    33      0 SN2010FZ    39312  9.11 
            22:12:55.0 - 22:14:15.0    34      0 SN2010FZ    44226  10   
            22:14:25.0 - 22:15:45.0    35      0 SN2010FZ    44226  10   
            22:15:55.0 - 22:17:15.0    36      0 SN2010FZ    44226  10   
            22:17:25.0 - 22:18:44.5    37      0 SN2010FZ    44226  9.89 
            22:18:54.0 - 22:20:14.0    38      0 SN2010FZ    44226  10   
            22:20:24.0 - 22:20:43.5    39      0 SN2010FZ    14742  9.6 
        (nVis = Total number of time/baseline visibilities per scan)
</pre>
(listing columns truncated) and we estimate about 37min on target. We had
about 25 antennas on average, and our spw selection picked out
610 channels (2MHz each) for a total of 1220MHz bandwidth.  If we plug this
into the
[https://science.nrao.edu/facilities/evla/calibration-and-tools/exposure "EVLA exposure calculator"]
then we find that we expect a rms thermal noise level of 7.6uJy so we are close!
 
[[Image:plotSN2010FZ_viewerfinal.png|200px|thumb|right|final image]]
Look at this in the viewer:
<source lang="python">
# In CASA
viewer('imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0')
</source>
Zoom in on the center (see figure to the right).
 
[[Image:viewSN2010FZ_spw0to11_mfs2tt1.png|200px|thumb|right|final tt1 image with box]]
In the previous section we demonstrated how to process and display the spectral index image. You can do
the same for this final image.  Here, we will do some rough analysis on the spectral index to determine
an intensity-weighted mean spectral index over the core region.
The <tt>.image.tt1</tt> from our mfs is an intensity times alpha image.  See the figure to the right.
Lets gate the Taylor-term images on intensity as before:
<source lang="python">
# In CASA
immath(imagename=['imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt1',
                  'imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0'],
      mode='evalexpr',
      expr='IM0[IM1>5.0E-5]',
      outfile='imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt1.filtered')
#
immath(imagename=['imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0'],
      mode='evalexpr',
      expr='IM0[IM0>5.0E-5]',
      outfile='imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0.filtered')
</source>
 
We can identify a box containing the central emission (see figure of tt1 in viewer) and note the corners.
(We could also use the region tools from the viewer, but that is for another exercise.)
Let us compute the intensity-weighted spectral index over this box by averaging
these masked images using {{imstat}} and computing the ratio:
<source lang="python">
# In CASA
mystat = imstat('imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt1.filtered',
                box='503,533,756,762')
avgtt0alpha = mystat['mean'][0]
#
mystat = imstat('imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0.filtered',
                box='503,533,756,762')
avgtt0 = mystat['mean'][0]
avgalpha = avgtt0alpha/avgtt0
print 'SN2010FZ I-weighted Alpha = '+str(avgalpha)
</source>
We get
<pre>
SN2010FZ I-weighted Alpha = -1.54687681754
</pre>
The emission in this source is on the steep side. At this point we do not know how reliable this is or
what we expect (though our calibrators come out with correct spectral indexes if we image them the
same way). But this illustrates a way to extract spectral information from our wideband mfs images.
 
As a final comparison, we turn to the Sloan Digital Sky Survey (SDSS) and a cutout image of our galaxy:
[[Image:NGC_2967_UGC_5180_IRAS_09394+0033_irg.jpg|400px|thumb|center|]]
from their [http://cosmo.nyu.edu/hogg/rc3/NGC_2967_UGC_5180_IRAS_09394+0033_irg.jpg RC3]
album (courtesy D.Hogg,M.Blanton,SDSS collaboration - see [[#Credits]]). This looks like a nice nearby
face-on spiral galaxy. How does our 6cm continuum emission line up with the optical?
 
Here is the EVLA 6cm image side by side with a i-band image from the Sloan Digital Sky Survey (SDSS) registered
to our image:
[[Image:plotSN2010FZ viewerfinalandSDSS.png|600px|thumb|center|final and sdss image]]
You can also find this image and load it into your viewer, and blink against our 6cm image.
 
We can also plot one as a raster and the other overlaid as contours. You can load the SDSS image
from the viewer Load Data panel and fiddle with contours. Once you know contour levels, you can
also use the imview task to load a raster and contour image:
<source lang="python">
# In CASA
imview(raster={ 'file' : 'imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0'},
      contour = { 'file' : 'NGC_2967_UGC_5180_IRAS_09394+0033-i.fits',
                  'levels' : [0.2, 0.5, 1, 1.5, 3],
                  'base' : 0.0,
                  'unit' : 1.0 } )
</source>
The figure below shows the SDSS contours overlaid on our 6cm image (after fiddling with the
colormap shift/slope for the EVLA raster image).
[[Image:viewSN2010FZ_spw0to11_mfs2tt0plusSDSS.png|400px|thumb|center|6cm EVLA raster plus SDSS i-band contours]]
 
Likewise, we can plot the SDSS image as a raster and overlay EVLA 6cm contours:
<source lang="python">
# In CASA
imview(raster={ 'file' : 'NGC_2967_UGC_5180_IRAS_09394+0033-i.fits',
                'scaling' : -2.0,
                'range' : [0,10] },
      contour = { 'file' : 'imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0',
                  'levels' : [0.04, 0.08, 0.16, 0.32, 0.64, 1.28, 2.56],
                  'base' : 0.0,
                  'unit' : 0.001 },
      zoom = { 'blc' : [397,300],
                'trc' : [1567,1231] } )
</source>
This is shown in the figure below.  Is the compact 6cm emission in upper left associated with a
spiral arm?
[[Image:viewSN2010FZ_spw0to11_SDSSiplusEVLA6cm.png|400px|thumb|center|SDSS i-band raster plus EVLA 6cm contours]]
 
== What to do next: some exercises for the user ==
 
Here are a number of things you can try after completing this tutorial:
 
# Use self-calibration to improve the data and re-clean to make a better image.
# Use multi-scale clean by adding non-zero scales to the <tt>multiscale</tt> parameter.
# Image the calibrators.  What sort of dynamic range can you get on them?  Is self-calibration needed (and if so what dynamic range do you get when you use it)?
# Try the <tt>testautoflag</tt> task (in 3.3.0 and later) to automatically flag RFI from the upper sideband.
 
== Credits ==
 
The EVLA data was taken by A. Soderberg et al. as part of project AS1015. See
[https://science.nrao.edu/enews/3.8/index.shtml#evlanoise NRAO eNews 3.8] (1-Sep-2010) for more on this result.
 
<blockquote><i>
The Expanded Very Large Array (EVLA) is a partnership of the United States, Canada, and Mexico. The EVLA is funded in the United States by the National Science Foundation, in Canada by the National Research Council, and in Mexico by the Comisión Nacional de Investigación Científica y Tecnológica (CONICyT).
</i></blockquote>
 
<blockquote><i>
The National Radio Astronomy Observatory is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc.
</i></blockquote>
 
SDSS image courtesy David Hogg & Michael Blanton, private communication.  Data comes from
SDSS DR7, see [http://adsabs.harvard.edu/abs/2009ApJS..182..543A Abazajian et. al 2009].
 
<blockquote><i>
Funding for the SDSS and SDSS-II has been provided by the Alfred P. Sloan Foundation, the Participating Institutions, the National Science Foundation, the U.S. Department of Energy, the National Aeronautics and Space Administration, the Japanese Monbukagakusho, the Max Planck Society, and the Higher Education Funding Council for England. The SDSS Web Site is [http://www.sdss.org/].
</i></blockquote>
 
<blockquote><i>
The SDSS is managed by the Astrophysical Research Consortium for the Participating Institutions. The Participating Institutions are the American Museum of Natural History, Astrophysical Institute Potsdam, University of Basel, University of Cambridge, Case Western Reserve University, University of Chicago, Drexel University, Fermilab, the Institute for Advanced Study, the Japan Participation Group, Johns Hopkins University, the Joint Institute for Nuclear Astrophysics, the Kavli Institute for Particle Astrophysics and Cosmology, the Korean Scientist Group, the Chinese Academy of Sciences (LAMOST), Los Alamos National Laboratory, the Max-Planck-Institute for Astronomy (MPIA), the Max-Planck-Institute for Astrophysics (MPA), New Mexico State University, Ohio State University, University of Pittsburgh, University of Portsmouth, Princeton University, the United States Naval Observatory, and the University of Washington.
</i></blockquote>

Latest revision as of 13:55, 12 November 2015

Redirect to:

Retrieved from "https://casaguides.nrao.edu/index.php?title=EVLA_6cmWideband_Tutorial_SN2010FZ&oldid=18207"