View source for VLA CASA Flagging-CASA4.5.2

* '''This CASA guide was designed for CASA 4.5.2''' 

== Overview ==

This CASA guide will cover a priori data flagging, including online flagging, shadowing, zero clipping, and quacking. It will also cover auto-flagging RFI (Radio Frequency Interference) via TFcrop (Time-Frequency crop) and rflag.

We will be utilizing data taken with the Karl G. Jansky, Very Large Array, of a supernova remnant [http://simbad.u-strasbg.fr/simbad/sim-id?Ident=SNR+G055.7%2B03.4&NbIdent=1&Radius=2&Radius.unit=arcmin&submit=submit+id G055.7+3.4.]. The data were taken on August 23, 2010, in the first D-configuration for which the new wide-band capabilities of the WIDAR (Wideband Interferometric Digital ARchitecture) correlator were available.  The 8-hour-long observation includes all available 1 GHz of bandwidth in L-band, from 1-2 GHz in frequency. The observations were made in 8-bit mode, which has the best dynamic range and is thus best suited for RFI excision.

The guide will often reference the CASA cookbook which is available [http://casa.nrao.edu/docs/cookbook/index.html online] and can be downloaded as a [http://casa.nrao.edu/casa_cookbook.pdf pdf]. 

== Obtaining the Data ==

A copy of the data <font color=green>(5.2GB)</font> can be downloaded from [http://casa.nrao.edu/Data/EVLA/SNRG55/SNR_G55_10s.tar.gz http://casa.nrao.edu/Data/EVLA/SNRG55/SNR_G55_10s.tar.gz]
These data are already in CASA MeasurementSet (MS) format and were prepared specifically for this guide. 

In the following we explain how the data were processed as it has implications for later flagging steps. First, the data were loaded from the [http://archive.nrao.edu NRAO archive]. On the archive search form, specify "Archive File ID" to "AB1345_sb1800808_1.55431.004049953706" and select "SDM-BDF dataset (all files)" as the data download option on the following page. But be aware that the full dataset has 170GB!

Once the SDM is downloaded, create a MeasurementSet with either task importevla or importasdm. Here we are using the second option:

'''importasdm(asdm='AB1345_sb1800808_1.55431.004049953706', vis='SNR_G55.ms', process_flags=True, tbuff=1.5, applyflags=False, outfile='SNR_G55.ms.onlineflags.txt', flagbackup=False)'''

    process_flags = True: This parameter controls the creation of online flags from the Flag.xml SDM table. It will create online flags in the FLAG_CMD sub-table within the MS (more on this later).
    tbuff = 1.5: This parameter adds a time "buffer" padding to the flags in both directions to deal with timing mismatches. This is important for JVLA data taken before April 2011. This value should be set to 1.5x integration time. This particular observation had 1 second integrations. (CASA Cookbook section 2.2.2 and 3.5.1.3)
    applyflags = False: We will apply these flags later in the tutorial.
    outfile = 'SNR_G55.ms.onlineflags.txt': This will create a text file with a list of online flags that can be applied. Created just for convenience, and won't be used during the tutorial.

Split and time-averaged using the split task. For CASA 4.5.3 we use the split2() implementation:

'''split2(vis='SNR_G55.ms', outputvis='SNR_G55_10s.ms', field='3~5', spw='4~5,7~8', antenna='!ea06,ea17,ea20,ea26', datacolumn='data', keepflags=False, timebin='10s')'''

    field: Several fields have been removed which we won't be utilizing. We are only keeping fields 3-5.
    spw: Several spectral windows have been removed which were polluted with RFI, which neither auto-flagging nor hand flagging could remedy. We are keeping spectral windows 4,5,7,8.
    antenna: Several antennas have been removed (!), which at the time of the observation, didn't have an L-Band receiver installed. (more on this later)
    Time-averaged to 10-seconds, to reduce size and processing time when running tasks.


== Unpack the Data == 

Once you've downloaded the file from the link above, let's untar it:
<source lang="bash">
tar -xzvf SNR_G55_10s.tar.gz
</source>

That will create a directory called SNR_G55_10s.ms <font color=green>(6.1GB)</font>, which is the MS.

== Starting CASA ==

Start CASA by typing '''casa''' on a terminal command line.  If you have not used CASA before, some helpful tips are available on the [[Getting Started in CASA]] page.

This guide has been written for CASA version 4.5.3.  Please confirm your version before proceeding by checking the message in the command line interface window or the CASA logger after startup.

For this tutorial, we will be running tasks using the ''task (parameter = value)'' syntax. When called in this manner, all parameters not explicitly set will use their default values. 

== Preliminary Data Evaluation == 

As a first step, use {{listobs}} to have a look at the MS:

<source lang="python">
# In CASA
listobs(vis='SNR_G55_10s.ms', listfile='SNR_G55_10s.listobs')
</source>

*listfile: Creates a text document "SNR_G55_10s.listobs" with details for the observation:
<pre>
================================================================================
           MeasurementSet Name:  /lustre/aoc/sciops/jott/casa/topicalguide/SNR_G55_10s.ms      MS Version 2
================================================================================
   Observer: Dr. Sanjay Sanjay Bhatnagar     Project: uid://evla/pdb/1072564  
Observation: EVLA
Data records: 3024000       Total elapsed time = 26926 seconds
   Observed from   23-Aug-2010/00:56:36.0   to   23-Aug-2010/08:25:22.0 (UTC)

   ObservationID = 0         ArrayID = 0
  Date        Timerange (UTC)          Scan  FldId FieldName             nRows     SpwIds   Average Interval(s)    ScanIntent
  23-Aug-2010/00:56:36.0 - 00:58:06.0    14      0 J1925+2106               10080  [0,1,2,3]  [10, 10, 10, 10] [CALIBRATE_PHASE#UNSPECIFIED]
              00:58:06.0 - 00:59:36.0    15      0 J1925+2106               10080  [0,1,2,3]  [10, 10, 10, 10] [CALIBRATE_PHASE#UNSPECIFIED]
              00:59:36.0 - 01:01:05.0    16      0 J1925+2106               10080  [0,1,2,3]  [9.89, 9.89, 9.89, 9.89] [CALIBRATE_PHASE#UNSPECIFIED]
              01:01:05.0 - 01:02:35.0    17      0 J1925+2106               10080  [0,1,2,3]  [10, 10, 10, 10] [CALIBRATE_PHASE#UNSPECIFIED]
              01:02:35.0 - 01:04:05.0    18      0 J1925+2106               10080  [0,1,2,3]  [10, 10, 10, 10] [CALIBRATE_PHASE#UNSPECIFIED]
              01:04:05.0 - 01:05:34.0    19      0 J1925+2106               10080  [0,1,2,3]  [9.89, 9.89, 9.89, 9.89] [CALIBRATE_PHASE#UNSPECIFIED]
              01:05:34.0 - 01:07:04.0    20      0 J1925+2106               10080  [0,1,2,3]  [10, 10, 10, 10] [CALIBRATE_PHASE#UNSPECIFIED]
              01:07:04.0 - 01:08:34.0    21      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              01:08:34.0 - 01:10:04.0    22      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              01:10:04.0 - 01:11:34.0    23      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              01:11:34.0 - 01:13:03.0    24      1 G55.7+3.4                10080  [0,1,2,3]  [9.89, 9.89, 9.89, 9.89] [OBSERVE_TARGET#UNSPECIFIED]
              01:13:03.0 - 01:14:33.0    25      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              01:14:33.0 - 01:16:03.0    26      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              01:16:03.0 - 01:17:33.0    27      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              01:17:33.0 - 01:19:02.0    28      1 G55.7+3.4                10080  [0,1,2,3]  [9.89, 9.89, 9.89, 9.89] [OBSERVE_TARGET#UNSPECIFIED]
              01:19:02.0 - 01:20:32.0    29      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              01:20:32.0 - 01:22:02.0    30      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              01:22:02.0 - 01:23:32.0    31      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              01:23:32.0 - 01:25:01.0    32      1 G55.7+3.4                10080  [0,1,2,3]  [9.89, 9.89, 9.89, 9.89] [OBSERVE_TARGET#UNSPECIFIED]
              01:25:01.0 - 01:26:31.0    33      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              01:26:31.0 - 01:28:01.0    34      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              01:28:01.0 - 01:29:31.0    35      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              01:29:31.0 - 01:31:00.0    36      1 G55.7+3.4                10080  [0,1,2,3]  [9.89, 9.89, 9.89, 9.89] [OBSERVE_TARGET#UNSPECIFIED]
              01:31:00.0 - 01:32:30.0    37      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              01:32:30.0 - 01:34:00.0    38      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              01:34:00.0 - 01:35:30.0    39      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              01:35:30.0 - 01:36:59.0    40      1 G55.7+3.4                10080  [0,1,2,3]  [9.89, 9.89, 9.89, 9.89] [OBSERVE_TARGET#UNSPECIFIED]
              01:36:59.0 - 01:38:29.0    41      0 J1925+2106               10080  [0,1,2,3]  [10, 10, 10, 10] [CALIBRATE_PHASE#UNSPECIFIED]
              01:38:29.0 - 01:39:59.0    42      0 J1925+2106               10080  [0,1,2,3]  [10, 10, 10, 10] [CALIBRATE_PHASE#UNSPECIFIED]
              01:39:59.0 - 01:41:29.0    43      0 J1925+2106               10080  [0,1,2,3]  [10, 10, 10, 10] [CALIBRATE_PHASE#UNSPECIFIED]
              01:41:29.0 - 01:42:58.0    44      1 G55.7+3.4                10080  [0,1,2,3]  [9.89, 9.89, 9.89, 9.89] [OBSERVE_TARGET#UNSPECIFIED]
<snip>
              08:11:54.0 - 08:13:24.0   305      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              08:13:24.0 - 08:14:54.0   306      1 G55.7+3.4                10080  [0,1,2,3]  [10, 10, 10, 10] [OBSERVE_TARGET#UNSPECIFIED]
              08:14:54.0 - 08:16:24.0   307      2 0542+498=3C147           10080  [0,1,2,3]  [10, 10, 10, 10] [CALIBRATE_AMPLI#UNSPECIFIED,CALIBRATE_BANDPASS#UNSPECIFIED,UNSPECIFIED#UNSPECIFIED]
              08:16:24.0 - 08:17:54.0   308      2 0542+498=3C147           10080  [0,1,2,3]  [10, 10, 10, 10] [CALIBRATE_AMPLI#UNSPECIFIED,CALIBRATE_BANDPASS#UNSPECIFIED,UNSPECIFIED#UNSPECIFIED]
              08:17:54.0 - 08:19:23.0   309      2 0542+498=3C147           10080  [0,1,2,3]  [9.89, 9.89, 9.89, 9.89] [CALIBRATE_AMPLI#UNSPECIFIED,CALIBRATE_BANDPASS#UNSPECIFIED,UNSPECIFIED#UNSPECIFIED]
              08:19:23.0 - 08:20:53.0   310      2 0542+498=3C147           10080  [0,1,2,3]  [10, 10, 10, 10] [CALIBRATE_AMPLI#UNSPECIFIED,CALIBRATE_BANDPASS#UNSPECIFIED,UNSPECIFIED#UNSPECIFIED]
              08:20:53.0 - 08:22:23.0   311      2 0542+498=3C147           10080  [0,1,2,3]  [10, 10, 10, 10] [CALIBRATE_AMPLI#UNSPECIFIED,CALIBRATE_BANDPASS#UNSPECIFIED,UNSPECIFIED#UNSPECIFIED]
              08:22:23.0 - 08:23:52.0   312      2 0542+498=3C147           10080  [0,1,2,3]  [9.89, 9.89, 9.89, 9.89] [CALIBRATE_AMPLI#UNSPECIFIED,CALIBRATE_BANDPASS#UNSPECIFIED,UNSPECIFIED#UNSPECIFIED]
              08:23:52.0 - 08:25:22.0   313      2 0542+498=3C147           10080  [0, 1, 2, 3]  [10, 10, 10, 10] [CALIBRATE_AMPLI#UNSPECIFIED,CALIBRATE_BANDPASS#UNSPECIFIED,UNSPECIFIED#UNSPECIFIED]
           (nRows = Total number of rows per scan) 
Fields: 3
  ID   Code Name                RA               Decl           Epoch   SrcId      nRows
  0    D    J1925+2106          19:25:59.605371 +21.06.26.16218 J2000   0         433440
  1    NONE G55.7+3.4           19:21:40.000000 +21.45.00.00000 J2000   1        2520000
  2    N    0542+498=3C147      05:42:36.137916 +49.51.07.23356 J2000   2          70560
Spectral Windows:  (4 unique spectral windows and 1 unique polarization setups)
  SpwID  Name      #Chans   Frame   Ch0(MHz)  ChanWid(kHz)  TotBW(kHz) CtrFreq(MHz) BBC Num  Corrs          
  0      Subband:0     64   TOPO    1256.000      2000.000    128000.0   1319.0000        4  RR  RL  LR  LL
  1      Subband:2     64   TOPO    1384.000      2000.000    128000.0   1447.0000        4  RR  RL  LR  LL
  2      Subband:1     64   TOPO    1648.000      2000.000    128000.0   1711.0000        8  RR  RL  LR  LL
  3      Subband:0     64   TOPO    1776.000      2000.000    128000.0   1839.0000        8  RR  RL  LR  LL
Sources: 12
  ID   Name                SpwId RestFreq(MHz)  SysVel(km/s) 
  0    J1925+2106          0     -              -            
  0    J1925+2106          1     -              -            
  0    J1925+2106          2     -              -            
  0    J1925+2106          3     -              -            
  1    G55.7+3.4           0     -              -            
  1    G55.7+3.4           1     -              -            
  1    G55.7+3.4           2     -              -            
  1    G55.7+3.4           3     -              -            
  2    0542+498=3C147      0     -              -            
  2    0542+498=3C147      1     -              -            
  2    0542+498=3C147      2     -              -            
  2    0542+498=3C147      3     -              -            
Antennas: 27:
  ID   Name  Station   Diam.    Long.         Lat.                Offset from array center (m)                ITRF Geocentric coordinates (m)        
                                                                     East         North     Elevation               x               y               z
  0    ea01  W09       25.0 m   -107.37.25.2  +33.53.51.0       -521.9416     -332.7766       -1.2001 -1601710.017000 -5042006.925200  3554602.355600
  1    ea02  E02       25.0 m   -107.37.04.4  +33.54.01.1          9.8240      -20.4293       -2.7806 -1601150.060300 -5042000.619800  3554860.729400
  2    ea03  E09       25.0 m   -107.36.45.1  +33.53.53.6        506.0564     -251.8670       -3.5825 -1600715.950800 -5042273.187000  3554668.184500
  3    ea04  W01       25.0 m   -107.37.05.9  +33.54.00.5        -27.3562      -41.3030       -2.7418 -1601189.030140 -5042000.493300  3554843.425700
  4    ea05  W08       25.0 m   -107.37.21.6  +33.53.53.0       -432.1167     -272.1478       -1.5054 -1601614.091000 -5042001.652900  3554652.509300
  5    ea06  N06       25.0 m   -107.37.06.9  +33.54.10.3        -54.0649      263.8778       -4.2273 -1601162.591000 -5041828.999000  3555095.896400
  6    ea07  E05       25.0 m   -107.36.58.4  +33.53.58.8        164.9788      -92.8032       -2.5268 -1601014.462000 -5042086.252000  3554800.799800
  7    ea08  N01       25.0 m   -107.37.06.0  +33.54.01.8        -30.8810       -1.4664       -2.8597 -1601185.634945 -5041978.156586  3554876.424700
  8    ea09  E06       25.0 m   -107.36.55.6  +33.53.57.7        236.9058     -126.3369       -2.4443 -1600951.588000 -5042125.911000  3554773.012300
  9    ea10  N03       25.0 m   -107.37.06.3  +33.54.04.8        -39.0773       93.0192       -3.3330 -1601177.376760 -5041925.073200  3554954.584100
  10   ea11  E04       25.0 m   -107.37.00.8  +33.53.59.7        102.8054      -63.7682       -2.6414 -1601068.790300 -5042051.910200  3554824.835300
  11   ea12  E08       25.0 m   -107.36.48.9  +33.53.55.1        407.8285     -206.0065       -3.2272 -1600801.926000 -5042219.366500  3554706.448200
  12   ea13  N07       25.0 m   -107.37.07.2  +33.54.12.9        -61.1037      344.2331       -4.6138 -1601155.635800 -5041783.843800  3555162.374100
  13   ea15  W06       25.0 m   -107.37.15.6  +33.53.56.4       -275.8288     -166.7451       -2.0590 -1601447.198000 -5041992.502500  3554739.687600
  14   ea16  W02       25.0 m   -107.37.07.5  +33.54.00.9        -67.9687      -26.5614       -2.7175 -1601225.255200 -5041980.383590  3554855.675000
  15   ea17  W07       25.0 m   -107.37.18.4  +33.53.54.8       -349.9877     -216.7509       -1.7975 -1601526.387300 -5041996.840100  3554698.327400
  16   ea18  N09       25.0 m   -107.37.07.8  +33.54.19.0        -77.4346      530.6273       -5.5859 -1601139.485100 -5041679.036800  3555316.533200
  17   ea19  W04       25.0 m   -107.37.10.8  +33.53.59.1       -152.8599      -83.8054       -2.4614 -1601315.893000 -5041985.320170  3554808.304600
  18   ea20  N05       25.0 m   -107.37.06.7  +33.54.08.0        -47.8454      192.6015       -3.8723 -1601168.786100 -5041869.054000  3555036.936000
  19   ea21  E01       25.0 m   -107.37.05.7  +33.53.59.2        -23.8638      -81.1510       -2.5851 -1601192.467800 -5042022.856800  3554810.438800
  20   ea22  N04       25.0 m   -107.37.06.5  +33.54.06.1        -42.6239      132.8436       -3.5494 -1601173.979400 -5041902.657700  3554987.517500
  21   ea23  E07       25.0 m   -107.36.52.4  +33.53.56.5        318.0509     -164.1850       -2.6957 -1600880.571400 -5042170.388000  3554741.457400
  22   ea24  W05       25.0 m   -107.37.13.0  +33.53.57.8       -210.0959     -122.3887       -2.2577 -1601377.009500 -5041988.665500  3554776.393400
  23   ea25  N02       25.0 m   -107.37.06.2  +33.54.03.5        -35.6245       53.1806       -3.1345 -1601180.861480 -5041947.453400  3554921.628700
  24   ea26  W03       25.0 m   -107.37.08.9  +33.54.00.1       -105.3447      -51.7177       -2.6037 -1601265.153600 -5041982.533050  3554834.858400
  25   ea27  E03       25.0 m   -107.37.02.8  +33.54.00.5         50.6641      -39.4835       -2.7273 -1601114.365500 -5042023.151800  3554844.944000
  26   ea28  N08       25.0 m   -107.37.07.5  +33.54.15.8        -68.9057      433.1889       -5.0602 -1601147.940400 -5041733.837000  3555235.956000

</pre>

* J1925+2106,     field ID 0: Phase Calibrator;
* G55.7+3.4,      field ID 1: The Supernova Remnant;
* 0542+498=3C147, field ID 2: Amplitude/Bandpass Calibrator

We can also see that these sources have associated "scan intents", which indicate their function in the observation.  Note that you can select sources based on their intents in certain CASA tasks.  The various scan intents in this data set are:

* CALIBRATE_PHASE indicates that this is a scan to be used for gain calibration;
* OBSERVE_TARGET indicates that this is the science target;
* CALIBRATE_AMPLI indicates that this is to be used for flux calibration; and 
* CALIBRATE_BANDPASS indicates that these scans are to be used for bandpass calibration.

Note that 3C147 is to be used for both flux and bandpass calibration.

It's important to also note that the antennas have a name and ID associated with them. For example antenna ID 15 is named ea17 ( The "ea" stemming from the Expanded VLA project). When specifying an antenna within a task parameter, we will mainly reference them by name. 

We can see the antenna configuration for this observation by using {{plotants}}:

[[Image:plotAnts.png|200px|thumb|right|plotants image showing the antenna configuration during the observation.]]
[[Image:amp_v_freq_rawdata.png|200px|thumb|right|plotms image of scan 190 showing strong RFI spikes in amplitude.]]

<source lang="python">
# In CASA
plotants(vis='SNR_G55_10s.ms', figfile='SNR_G55_10s.plotants.png')
</source>

This shows that antennas ea01, ea03, and ea18 were on the extreme ends of the west, east, and north arms, respectively.  The antenna position diagram is particularly useful as a guide to help determine which antenna to use as the reference antenna later during calibration. Note that antennas on stations 8 of each arm (N08, E08, W08) do not get moved during array reconfigurations, they can therefore at times be good choices as reference antennas. In this case, we'll probably want to choose something closer to the center of the array, with no shadowing. 

We may also inspect the raw data using {{plotms}}. To start with, lets look at a subset of scans on the supernova remnant:

<source lang="python">
# In CASA
plotms(vis='SNR_G55_10s.ms', scan='30,75,120,165,190,235,303', antenna='ea24', xaxis='freq', 
       yaxis='amp', coloraxis='spw', iteraxis='scan', correlation='RR,LL')
</source>

* coloraxis='spw': Parameter indicates that a different color will be assigned to each spectral window.
* antenna='ea24' : We chose only information for antenna ea24.
* iteraxis='scan': Parameter tells plotms to display a new plot for each scan. 
* correlation='RR,LL': We just want to display the right and left circular polarizations, without the cross-hand terms. 

Flipping through to scan 190, we can see that there is significant time and frequency variable RFI present in the observation, as seen by the large spikes in amplitude. In particular, we can see that several spectral windows are quite badly affected. To determine which spectral windows they are, click on the "Mark Regions" tool at the bottom of the plotms GUI (the open box with a green "plus" sign). Use the mouse to select a few of the highest-amplitude points in each of the spectral windows. Click on the "Locate" button (magnifying glass). Information about the selected areas should now display in the logger window. For example:
<br />
<br />
<pre>
Frequency in [1.30662 1.31407] or [1.32377 1.33569] or [1.67866 1.69581], Amp in [0.16871 0.217097] or [0.153226 0.186613] or [0.172581 0.235968]:
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:19:58.0 BL=ea16@W02 & ea24@W05[14&22] Spw=0 Chan=27 Freq=1.31  Corr=RR X=1.31  Y=0.17187   (1718/144/1718)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:19:58.0 BL=ea19@W04 & ea24@W05[17&22] Spw=0 Chan=27 Freq=1.31  Corr=LL X=1.31  Y=0.178607  (1975/144/1975)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:19:58.0 BL=ea22@N04 & ea24@W05[20&22] Spw=0 Chan=37 Freq=1.33  Corr=RR X=1.33  Y=0.18105   (2250/144/2250)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:08.0 BL=ea19@W04 & ea24@W05[17&22] Spw=0 Chan=27 Freq=1.31  Corr=LL X=1.31  Y=0.21228   (1975/145/1975)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:08.0 BL=ea21@E01 & ea24@W05[19&22] Spw=0 Chan=27 Freq=1.31  Corr=LL X=1.31  Y=0.181632  (2103/145/2103)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:08.0 BL=ea22@N04 & ea24@W05[20&22] Spw=0 Chan=37 Freq=1.33  Corr=RR X=1.33  Y=0.167068  (2250/145/2250)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:28.0 BL=ea19@W04 & ea24@W05[17&22] Spw=0 Chan=27 Freq=1.31  Corr=LL X=1.31  Y=0.216252  (1818323722/147/1975)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:28.0 BL=ea22@N04 & ea24@W05[20&22] Spw=0 Chan=37 Freq=1.33  Corr=RR X=1.33  Y=0.182341  (1818323997/147/2250)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:38.0 BL=ea07@E05 & ea24@W05[6&22]  Spw=0 Chan=27 Freq=1.31  Corr=LL X=1.31  Y=0.175032  (1866691864/148/695)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:38.0 BL=ea09@E06 & ea24@W05[8&22]  Spw=0 Chan=27 Freq=1.31  Corr=RR X=1.31  Y=0.176728  (1866692119/148/950)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:38.0 BL=ea16@W02 & ea24@W05[14&22] Spw=0 Chan=27 Freq=1.31  Corr=LL X=1.31  Y=0.170435  (1866692888/148/1719)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:38.0 BL=ea19@W04 & ea24@W05[17&22] Spw=0 Chan=27 Freq=1.31  Corr=LL X=1.31  Y=0.210123  (1866693144/148/1975)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:38.0 BL=ea21@E01 & ea24@W05[19&22] Spw=0 Chan=27 Freq=1.31  Corr=LL X=1.31  Y=0.185065  (1866693272/148/2103)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:38.0 BL=ea22@N04 & ea24@W05[20&22] Spw=0 Chan=37 Freq=1.33  Corr=RR X=1.33  Y=0.158359  (1866693419/148/2250)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:38.0 BL=ea24@W05 & ea27@E03[22&25] Spw=0 Chan=27 Freq=1.31  Corr=RR X=1.31  Y=0.190659  (1866693783/148/2614)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:48.0 BL=ea21@E01 & ea24@W05[19&22] Spw=0 Chan=27 Freq=1.31  Corr=LL X=1.31  Y=0.197052  (1852669347/149/2103)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:48.0 BL=ea22@N04 & ea24@W05[20&22] Spw=0 Chan=37 Freq=1.33  Corr=RR X=1.33  Y=0.183905  (1852669494/149/2250)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:48.0 BL=ea24@W05 & ea27@E03[22&25] Spw=0 Chan=27 Freq=1.31  Corr=RR X=1.31  Y=0.170542  (1852669858/149/2614)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:58.0 BL=ea21@E01 & ea24@W05[19&22] Spw=0 Chan=27 Freq=1.31  Corr=LL X=1.31  Y=0.179017  (1668509051/150/2103)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:58.0 BL=ea22@N04 & ea24@W05[20&22] Spw=0 Chan=37 Freq=1.33  Corr=RR X=1.33  Y=0.177841  (1668509198/150/2250)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:21:08.0 BL=ea21@E01 & ea24@W05[19&22] Spw=0 Chan=27 Freq=1.31  Corr=LL X=1.31  Y=0.193158  (1866681459/151/2103)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:19:58.0 BL=ea19@W04 & ea24@W05[17&22] Spw=2 Chan=19 Freq=1.686 Corr=RR X=1.686 Y=0.209811  (4844192/162/1958)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:19:58.0 BL=ea19@W04 & ea24@W05[17&22] Spw=2 Chan=19 Freq=1.686 Corr=LL X=1.686 Y=0.224424  (4844193/162/1959)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:08.0 BL=ea19@W04 & ea24@W05[17&22] Spw=2 Chan=19 Freq=1.686 Corr=RR X=1.686 Y=0.231918  (1374179750/163/1958)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:08.0 BL=ea19@W04 & ea24@W05[17&22] Spw=2 Chan=19 Freq=1.686 Corr=LL X=1.686 Y=0.209708  (1374179751/163/1959)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:18.0 BL=ea08@N01 & ea24@W05[7&22]  Spw=2 Chan=19 Freq=1.686 Corr=RR X=1.686 Y=0.17744   (2898499/164/806)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:18.0 BL=ea10@N03 & ea24@W05[9&22]  Spw=2 Chan=19 Freq=1.686 Corr=LL X=1.686 Y=0.177235  (2898756/164/1063)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:18.0 BL=ea19@W04 & ea24@W05[17&22] Spw=2 Chan=19 Freq=1.686 Corr=RR X=1.686 Y=0.230346  (2899651/164/1958)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:18.0 BL=ea19@W04 & ea24@W05[17&22] Spw=2 Chan=19 Freq=1.686 Corr=LL X=1.686 Y=0.199502  (2899652/164/1959)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:28.0 BL=ea08@N01 & ea24@W05[7&22]  Spw=2 Chan=19 Freq=1.686 Corr=RR X=1.686 Y=0.177315  (6083110/165/806)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:28.0 BL=ea10@N03 & ea24@W05[9&22]  Spw=2 Chan=19 Freq=1.686 Corr=LL X=1.686 Y=0.17881   (6083367/165/1063)
Scan=190 Field=G55.7+3.4[1] Time=2010/08/23/05:20:28.0 BL=ea19@W04 & ea24@W05[17&22] Spw=2 Chan=19 Freq=1.686 Corr=RR X=1.686 Y=0.217573  (6084262/165/1958)
Found 45 points (45 unflagged) among 101376 in 0s.
</pre>

[[Image:Lband_RFI.png|400px|thumb|right|Side-by-side plots of RFI within a portion of L-Band and spectral window 0 of the observation.]]

We can see that Spw 0 and 2, are the worst affected by RFI. Also, we get the corresponding frequency where the RFI is present, which appears to be 1.31 and 1.33GHz for spectral window 0, and 1.686GHz for spectral window 2. 

The VLA has obtained RFI scans over its frequency range. And the results are presented on the [https://science.nrao.edu/facilities/vla/docs/manuals/obsguide/modes/rfi Radio Frequency Interference website.] Most of the RFI features can be identified, e.g. as radar, communications, satellites, airplanes, or birdies, i.e. RFI generated by the VLA electronics itself. 
For our observations, we can inspect the L-band specific RFI plots [https://science.nrao.edu/facilities/vla/observing/RFI/L-Band website]. We find that the 1310 and 1330 MHz features are due to FAA ASR radars, and the one at 1686MHz is most likely due to a GOES weather satellite. We can compare the plots of the FAA ASR radar from the website, with that of spw 0, and see that the spikes are practically identical and fall within the same frequencies. 

== Identifying Problematic Antennas from the Operator Logs ==

We first check the operator log for the observation to see if there were any issues noted during the run that need to be addressed. The logs are available [http://www.vla.nrao.edu/cgi-bin/oplogs.cgi on this website] and the observing log file for our observation can be found [http://www.vla.nrao.edu/operators/logs/2010/8/2010-08-23_0005_AB1345.pdf here].

The log has various pieces of information, including the start/end times for the observation, frequency bands used, weather, baseline information for recently moved antennas, and any outages or issues that may have been encountered. We can see that antenna ea07 may need position corrections (see one of our calibration tutorials, e.g. [https://casaguides.nrao.edu/index.php?title=EVLA_high_frequency_Spectral_Line_tutorial_-_IRC%2B10216 IRC+10216] on how to fix this), and several antennas are missing an L-Band receiver, including ea06, ea17, ea20, and ea26. These antennas were already removed from the dataset used here (as explained above in "Obtaining the Data") and we can continue with online flagging.

== Online Flags ==

At the time of importing from the SDM-BDF raw data to a MS, we chose to process the online flags from the Flags.xml file to the FLAG_CMD sub-table within the MS. We also created a txt document which includes a list of online flags. 

The Flags.xml file holds information of flags created during the observation, such as subreflector issues, and antennas not being on source. 
We will now want to apply these online flags to the data, but first, let's create a plot of the flags we are about to apply to get an idea of what will be flagged.

[[Image:flaggingreason_vs_time.png|300px|thumb|right|Online Flags]]

We will employ the {{flagcmd}} task to apply the online flags. 

<source lang="python">
# In CASA
flagcmd(vis='SNR_G55_10s.ms', inpmode='table', reason='any', action='plot', plotfile='flaggingreason_vs_time.png')
</source>

We can see several instances of online flagging in the created image. Most notably, ea28 and ea08 had some subreflector issues througout the observation. Online flags are instances of possible missing data, including:

* ANTENNA_NOT_ON_SOURCE 
The VLA antennas have slewing speeds of 20 degrees per minute in azimuth, and 40 degrees per minute in elevation. Some antennas are slower than others, and may take a few more seconds to reach the next source. The antennas can also take a few seconds to settle down due to small oscillations after having slewed. 

* SUBREFLECTOR_ERROR    
The FRM (Focus Rotation Mount) located at the apex of the antennas, is responsible for focusing the incoming radio signal to the corresponding receiver. They can at times have issues with their focus and/or rotation axes. Being off target, so much as a few fractions of a degree can result in loss of data, depending on the frequency being observed.

Now that we've plotted the online flags, we will apply them to the MS. 

<source lang="python">
# In CASA
flagcmd(vis='SNR_G55_10s.ms', inpmode='table', reason='any', action='apply', flagbackup=False)
</source>

The CASA logger should report the progress as the task applies these flags in chunks. Once it has finished, it will report on the percentage of data that has been flagged. 

=== Shadowed Antennas ===

Since this is the most compact VLA configuration, there may be instances where one antenna blocks, or "shadows" another.  {{flagdata}} (CASA Cookbook 3.4) can determine shadowed antennas by their location and observation direction:

<source lang="python">
# In CASA
flagdata(vis='SNR_G55_10s.ms', mode='shadow', tolerance=0.0, flagbackup=False)
</source>

In this particular observation, there does not appear to be much data affected by shadowing, as can be seen in the logger report. One reason why this may be the case, is the antennas were pointed at sources high in elevation.

=== Zero-Amplitude Data ===

In addition to shadowing, there may be times during which the correlator writes out pure zero-valued data.  In order to remove this bad data, we run {{flagdata}} to remove any pure zeroes:

<source lang="python">
# In CASA
flagdata(vis='SNR_G55_10s.ms', mode='clip', clipzeros=True, flagbackup=False) 
</source>

Inspecting the logger output which is generated by {{flagdata}} shows that there is a small quantity of zero-valued data (4.6%) present in this MS.

=== Quacking ===
Now we can utilize the flagdata task one more time in order to run it in quaking mode.

It's common for the array to "settle down" at the start of a scan. Quacking is used to remove data at scan boundaries, and it can apply the same edit to all scans for all baselines. 
 
<source lang="python">
# In CAS
flagdata(vis='SNR_G55_10s.ms', mode='quack', quackinterval=5.0, quackmode='beg', flagbackup=False)
</source>

*quackmode='beg'  : Data from the start of each scan will be flagged.
*quackinterval=5.0: Flag the first 5 seconds of every scan.

=== Backup Data - Flagmanager ===

Flags can be backed up in a file MS.flagversions that is related to the MS. For our MS the related flag backups are stored in 'SNR_G55_10s.ms.flagversions'. 

''Note: flags that are applied to the data are contained in the MS. MS.flagversions only contains backup flags that can be restored if required.''

Most flagging tasks, like {{flagdata}} offer a ''flagbackup'' parameter, that controls whether or not the current flags are being backed up in MS.flagversions before new (additional) flags will be applied. 

Backup flags can also be manually saved to, or restored from MS.flagversions using the {{flagmanager}} task.   

Now that we've applied online flags, clipped zero amplitude data, and removed shadowed data, we will create a backup of the flags in such a way, giving it the label 'after_priori_flagging': 

<source lang="python">
# In CASA
flagmanager(vis='SNR_G55_10s.ms', mode='save', versionname='after_priori_flagging')
</source>

From here on forward, if we make a mistake, we can always revert back to this version of the MS by setting mode='restore', and providing the version name we want to restore back to.

=== Hanning-Smoothing ===

Strong RFI sources can give rise to the Gibbs phenomenon. This is seen by 'ringing', a zig-zag pattern across the channels that neighbor the strong, usually narrow RFI. To remedy this ringing across the frequency channels, we can employ the Hanning smoothing algorithm via the 'hanningsmooth2' task (an implementation based on {{mstransform}}. Hanning smoothing applies a triangle kernel across the pattern which reduces the ringing and thus reduces the number of channels that are affected by RFI. This smoothing procedure, however, will also decrease the spectral resolution by a factor of two.

Note that running this task overwrites the data, and it cannot be reversed. Also, there is a loss in spectral resolution by a factor of two. In addition, the task allows for the creation of a new MS, or to directly operate on the requested column. For this tutorial, we will be directly operating on the data column. Note that hanning-smoothing will remove amplitude spikes, it is therefore not recommended for spectral analysis related science, such as HI. Also note that hanning smoothing cannot be reversed once you apply it to your data.

Let's create before and after images with the {{plotms}} task, to see the effects of hanning-smoothing the data. 

[[Image:amp_v_freq_before_after_hanning.gif|250px|thumb|right|The effects of applying Hanning-Smoothing to our data. Click on image to see a GIF movie of the comparison.]]

<source lang="python">
# In CASA
plotms(vis='SNR_G55_10s.ms', scan='190', antenna='ea24', spw='0~2',
    xaxis='freq', yaxis='amp', coloraxis='spw', title='Before Hanning',
    correlation='RR,LL', plotrange=[1.2,1.8,-0.01,0.25], 
    plotfile='amp_v_freq.beforeHanning.png')

hanningsmooth2(vis='SNR_G55_10s.ms', outputvis='SNR_G55_10s-hanning.ms', datacolumn='data')

plotms(vis='SNR_G55_10s-hanning.ms', scan='190', antenna='ea24', spw='0~2', 
    xaxis='freq', yaxis='amp', coloraxis='spw', title='After Hanning',
    correlation='RR,LL', plotrange=[1.2,1.8,-0.01,0.25], 
    plotfile='amp_v_freq.afterHanning.png')
</source>

<!--
Note that the second call to plotms includes a spectral window outside of the plot range. This is to force plotms to reload the plot. By default, plotms will not redraw a plot if the inputs are unchanged. Within the GUI, checking the "reload" box next to the "Plot" button will do this as well. 
-->

We can compare the generated plots and take notice that single channel RFI has spread into three channels, but it has also removed some of the worst RFI. This spreading of RFI to other channels will ultimately result in a little more data being flagged.

== Automatic RFI excision ==

Now that we're done with priori flagging, we can move on to removing some of the RFI present with auto-flagging algorithms used within {{flagdata}}. We will employ two CASA tasks, {{tfcrop}} and {{rflag}}. For further details on the two algorithm based tasks we will employ, please see the [https://science.nrao.edu/science/meetings/2016/vla-data-reduction/EVLAWorkshop_RFI_2016_UR.pdf presentation] given at the 5th VLA Data Reduction Workshop by Urvashi Rau. 

=== TFcrop ===

{{tfcrop}} is an algorithm that detects outliers in the 2D time-frequency plane, and can operate on un-calibrated data (non bandpass-corrected). Tfcrop will iterate through chunks of time, and undergo several steps in order to find and excise different types of RFI. (CASA Cookbook 3.4.2.7)

Step 1: Detect short-duration RFI spikes (narrow-band and broad-band). <br />
Step 2: Search for time-persistent RFI. <br />
Step 3: Search for time-persistent, narrow-band RFI. <br />
Step 4: Search for low-level wings of very strong RFI. <br />

More details on the algorithm steps can be found on this [http://www.aoc.nrao.edu/~rurvashi/TFCrop/TFCropV1/TFCropV1.html webpage.]

We will apply the auto-flagging tfcrop algorithm for each spectral window. We will walk through the first spw and then include the rest with the same command. Let's first plot the corrected data for scan 190, with all spectral windows, so we can compare the data before and after tfcrop.

<source lang="python">
# In CASA
plotms(vis='SNR_G55_10s-hanning.ms', scan='190', antenna='ea24', xaxis='freq', iteraxis='scan', yaxis='amp', 
       ydatacolumn='data', plotfile='amp_v_freq_before_tfcrop.png', title='Before TFcrop',
       correlation='RR,LL', coloraxis='spw', plotrange=[1.2,2,-0.01,0.25])
</source>

Following are a set of {{flagdata}} commands which have been found to work reasonably well with these data. Please take some time to play with the parameters and the plotting capabilities.  Since these runs set display='both' and action='calculate', the flags are displayed but not actually written to the MS.  This allows one to try different sets of parameters before actually applying the flags to the data.

Some representative plots are also displayed.  Each column displays an individual polarization product; since we're using all four polarizations, from left to right are RR, RL, LR, and LL. The first row shows the data with current flags applied, and the second includes the flags generated by flagdata. The x-axis is channel number (the spectral window ID is displayed in the top title) and the y-axis of the first two rows is all integrations included in a time "chunk", set by the ntime parameter. These are the data considered by the tfcrop algorithm during its flagging process, and changes in ntime will have some (relatively small) affect on what data are flagged. 

Each plot page displays data for a single baseline and time chunk.  The buttons at the bottom allow one to step through baseline (backward as well as forward), spw, scan, and field; "Stop Display" will continue the flagging operation without the GUI, and "Quit" aborts the run.

''' spw 0 '''

<source lang="python">
# In CASA
flagdata(vis='SNR_G55_10s-hanning.ms', mode='tfcrop', spw='0', 
         datacolumn='data', action='calculate', 
         display='both', flagbackup=False)
</source>

[[Image:Tfcrop.png|250px|thumb|right|TFcrop flagging results. Top row before, bottom row, after tfcrop.]]

As we iterate through the different scans and baselines, we can see the flagging we can apply, represented by the blue areas. Let's now apply these flags by changing the action parameter.

<source lang="python">
# In CASA
flagdata(vis='SNR_G55_10s-hanning.ms', mode='tfcrop', spw='0', 
         datacolumn='data', action='apply', 
         display='', flagbackup=False)
</source>

The logger will report the percentage of flagged data in the table selection. We can now apply the tfcrop algorithm to the remaining spectral windows. 

[[Image:amp_v_freq_before_after_tfcrop_scan190.gif|250px|thumb|right|The effects of applying tfcrop to our data, for scan 190. Click image for a comparison of the data before and after tfcrop.]]

''' spw 1, 2, 3 '''
<source lang="python">
# In CASA
flagdata(vis='SNR_G55_10s-hanning.ms', mode='tfcrop', spw='1~3', 
         datacolumn='data', action='apply', 
         display='both', flagbackup=False)
</source>

After tfcrop has gone through and flagged some of the worst RFI, we can inspect the log report and take note of how much has been flagged.

We can also use {{plotms}} to review the effects of using tfcrop and compare the image to the one created before applying tfcrop. We can see great improvements, especially for spectral window 0 and 2, which had some of the worst RFI.

<source lang="python">
# In CASA
plotms(vis='SNR_G55_10s-hanning.ms', scan='190', antenna='ea24', xaxis='freq', iteraxis='scan', yaxis='amp', 
       ydatacolumn='data', plotfile='amp_v_freq_after_tfcrop.png', title='After TFcrop',
       correlation='RR,LL', coloraxis='spw', plotrange=[1.2,2,-0.01,0.25])
</source>

We now calculate the amount of flagged data so far in our Measurement Set by using the flagdata task with mode='summary', and assign the returned python dictionary to the variable 'flagInfo'. We then parse the dictionary to print some of the flagging statistics.

<source lang="python">
# In CASA
flagInfo = flagdata(vis='SNR_G55_10s-hanning.ms', mode='summary')

print("\n %2.1f%% of G55.7+3.4, %2.1f%% of 3C147, and %2.1f%% of J1925+2106 are flagged. \n" % 
     (100.0 * flagInfo['field']['G55.7+3.4']['flagged'] / flagInfo['field']['G55.7+3.4']['total'], 
     100.0 * flagInfo['field']['0542+498=3C147']['flagged'] / flagInfo['field']['0542+498=3C147']['total'], 
     100.0 * flagInfo['field']['J1925+2106']['flagged'] / flagInfo['field']['J1925+2106']['total']))

print("Spectral windows are flagged as follows:")

for spw in range(0,4):
     print("SPW %s: %2.1f%%" % (spw, 100.0 * flagInfo['spw'][str(spw)]['flagged'] / flagInfo['spw'][str(spw)]['total']))
</source>

The results indicate we have flagged a little over 22% of G55.7+3.4. We can now move on to the other auto-flagging algorithm, rflag.

=== RFlag ===

In order to get the best possible result from the automatic RFI excision with {{flagdata}}'s mode 'rflag', we will first apply bandpass calibration to the MS.  Since the RFI is time-variable, using the phase calibration source to make an average bandpass over the entire observation will mitigate the amount of RFI present in the calculated bandpass.  (For the final calibration, we will use the designated bandpass source 3C147; however, since this object was only observed in the last set of scans, it doesn't sample the time variability and would not provide a good average bandpass.)

Since there are likely to be gain variations over the course of the observation, we will run {{gaincal}} to solve for an initial set of antenna-based phases over a narrow range of channels.  Those solutions will be applied to the data when the bandpass solutions are determined.  While amplitude variations will have little effect on the bandpass solutions, it is important to solve for these phase variations with sufficient time resolution to prevent decorrelation when vector averaging the data in computing the bandpass solutions.

In order to choose a narrow range of channels for each spectral window which are relatively RFI-free over the course of the observation, we can look at the data with {{plotms}}.  Note that it's important to only solve for phase using a narrow channel range, since an antenna-specific delay will cause the phase to vary with respect to frequency over the spectral window, perhaps by a substantial amount.

<source lang="python">
# In CASA
plotms(vis='SNR_G55_10s-hanning.ms', scan='42,65,88,11,134,157', antenna='ea24', 
       xaxis='channel', yaxis='amp', iteraxis='spw', yselfscale=True, correlation='RR,LL')
</source>

* yselfscale=True: sets the y-scaling to be for the currently displayed spectral window, since some spectral windows have much worse RFI and will skew the scale for others.

Looking at these plots, we can choose appropriate channel ranges for each SPW:

<pre>
SPW 0: 20-24
SPW 1: 49-52
SPW 2: 38-41
SPW 3: 41-44
</pre>

Using these channel ranges, we run {{gaincal}} to calculate phase-only solutions that will be used as input during our initial bandpass calibration.  Remember - the calibration tables we are creating now are so that we can use an automatic RFI flagging algorithm. Our final calibration tables will be generated later, after automated flagging. Here are the inputs for our initial pre-bandpass phase calibration:

<source lang="python">
# In CASA
gaincal(vis='SNR_G55_10s-hanning.ms', caltable='SNR_G55_10s-hanning.initPh', field='J1925+2106', solint='int',
        spw='0:20~24,1:49~52,2:38~41,3:41~44', refant='ea24', minblperant=3,
        minsnr=3.0, calmode='p')
</source>

* caltable='SNR_G55_10s-hanning.initPh': this is the output calibration table that will be written.
* field='J1925+2106': this is the phase calibrator we will use to calibrate the phases.
* solint='int': we request a solution for each 10-second integration.
* pw='0:20~24,1:49~52,2:38~41,3:41~44': note the syntax of this selection: a ":" is used to separate the SPW from channel selection, and "~" is used to indicate an inclusive range.  
* refant='ea24': we have chosen ea24 as the reference antenna after inspecting the antenna position diagram (see above).  It is relatively close to, but not directly in, the center of the array, which could be important in D-configuration, since you don't want the reference antenna to have a high probability of being shadowed by nearby antennas.
* minblperant=3: the minimum number of baselines which must be present to attempt a phase solution.
* minsnr=3.0: the minimum signal-to-noise a solution must have to be considered acceptable.  Note that solutions which fail this test will cause these data to be flagged downstream of this calibration step.
* calmode='p': perform phase-only solutions.

Note that a number of solutions do not pass the requirements of the minimum 3 baselines (generating the terminal message "Insufficient unflagged antennas to proceed with this solve.") or minimum signal-to-noise ratio (outputting "n of x solutions rejected due to SNR being less than 3 ..."). The logger output indicates 324 (320 for spw0) solutions succeeded, out of 387 attempted. 

[[Image:gain.phase_v_time.plotcal.png|250px|thumb|right| Gain Phase vs. Time for spw0, for several antennas.]]

It's always a good idea to inspect the resulting calibration table with {{plotcal}}. We will first iterate over antennas for spectral window 0.

<source lang="python">
# In CASA
plotcal(caltable='SNR_G55_10s-hanning.initPh', xaxis='time', yaxis='phase', iteration='antenna', 
        spw='0', plotrange=[-1,-1,-180,180])
</source>

We can see that the phases do not change much over the course of the observation. Let's now inspect the other spectral windows for a few of the antennas. 

<source lang="python">
# In CASA
plotcal(caltable='SNR_G55_10s-hanning.initPh', xaxis='time', yaxis='phase', iteration='spw', 
        antenna='ea01,ea05,ea10,ea24', plotrange=[-1,-1,-180,180])
</source>

Now that we've determined our plots look fairly reasonable, we will create a time-averaged bandpass solutions for the phase calibration source using the {{bandpass}} task.

<source lang="python">
# In CASA
bandpass(vis='SNR_G55_10s-hanning.ms', caltable='SNR_G55_10s-hanning.initBP', field='J1925+2106', solint='inf', combine='scan', 
         refant='ea24', minblperant=3, minsnr=10.0, gaintable='SNR_G55_10s-hanning.initPh',
         interp='nearest', solnorm=False)
</source>

* solint='inf', combine='scan': the solution interval of 'inf' will automatically break by scans; this requests that the solution intervals be combined over scans, so that we will get one solution per antenna.
* gaintable= 'SNR_G55_10s-hanning.initPh': we will pre-apply the initial phase solutions.
* interp='nearest': by default, {{bandpass}} will use linear interpolation for pre-applied calibration.  However, we want the <i>nearest</i> phase solution to be used for a given time.

Again, we can see that a number of solutions have been rejected by our choice of <tt>minsnr</tt>.

[[Image:GainAmp_vs_Freq_bandpass.png|250px|thumb|right|Bandpasses for antennas ea01 - ea09]]
[[Image:plotcal_ea01ea05_interactive_flagging.png|250px|thumb|right|Interactive plotcal flagging of the offset points for ea01 and ea05.]]

Let us now inspect the resulting bandpass plots with plotcal. 

<source lang="python">
# In CASA
plotcal(caltable='SNR_G55_10s-hanning.initBP', xaxis='freq', yaxis='amp',
        iteration='antenna', subplot=331)
</source>

* subplot=331: displays 3x3 plots per screen

We notice that antenna's ea01 and ea05 have a point that is offset from the rest. Let's plot just these two antennas, locate the point on the plot, and flag it interactively through plotcal. Please note that interactive flagging within plotcal will not create a backup, therefore it may be wise to use {{flagmanager}} before doing so. 

<source lang="python">
# In CASA
plotcal(caltable='SNR_G55_10s-hanning.initBP', antenna='ea01,ea05', xaxis='freq', yaxis='amp',
        iteration='antenna', subplot=211)
</source>

We will highlight the points by clicking on the "Mark Region" button, drawing boxes over the points, and clicking on the "Flag" button. Before doing this, one could also get information on the points by clicking on the "Locate" button, which will display details about the highlighted regions. After having flagged the offset points, your plots should update.

We will now apply the bandpass calibration table using {{applycal}}.

<source lang="python">
# In CASA
applycal(vis='SNR_G55_10s-hanning.ms', gaintable='SNR_G55_10s-hanning.initBP', calwt=False)
</source>

This operation will flag data that correspond to flagged solutions, so {{applycal}} makes a backup version of the flags prior to operating on the data. Note that running {{applycal}} might take a little while.

Now that we have bandpass-corrected data with some RFI flagged out, we will run {{flagdata}} in rflag mode (CASA Cookbook 3.4.2.8). RFlag, like TFcrop, is an autoflag algorithm which uses a sliding window statistical filter. Data is iterated through in chunks of time, where statistics are accumulated, and thresholds calculated. 

Additional information on the algorithm used in rflag, as well as the statistical details that are undertaken can be found [http://www.aoc.nrao.edu/~rurvashi/FlaggerDocs/node5.html on this webpage] (sections 2.1.7).

We will use flagdata with mode='rflag', and action='calculate' to first review the amount of data to be flagged. We will also change the datacolumn parameter to 'corrected', since we've applied the bandpass corrections to the MS, and in the process, created the ''corrected_data'' column in the MS table. 

''' spw 0 '''

First, let's create a plot of our corrected data before we apply RFlag.

<source lang="python">
# In CASA
plotms(vis='SNR_G55_10s-hanning.ms', scan='190' , antenna='ea24',
xaxis='freq', yaxis='amp', ydatacolumn='corrected', coloraxis='spw', title='Before RFlag',
correlation='RR,LL', plotfile='amp_v_freq.beforeRFlag.png')

flagdata(vis='SNR_G55_10s-hanning.ms', mode='rflag', spw='0', datacolumn='corrected', 
         action='calculate', display='both', flagbackup=False)
</source>

[[Image:rflag_default_values.png|200px|thumb|left|RFlag with Default Parameters. (use the "next scan", "next baseline", and "next source" buttons to see this specific screen.) ]]
[[Image:rflag_cutoff_1.5_sigma.png|200px|thumb|right|RFlag with a cutoff of 1.5 Sigma, shown in blue. (use the "next scan", "next baseline", and "next source" buttons to see this specific screen.)]]

We can see that a lot of RFI is being missed by the default timedevscale and freqdevscale parameters of 5.0. Since we know spw 0 has lots of RFI, we will change the values to 1.5 and apply them. Note that we can always decide that this still isn't good enough and click on "Quit" to change our parameter values. 

<source lang="python">
# In CASA
flagdata(vis='SNR_G55_10s-hanning.ms', mode='rflag', spw='0', datacolumn='corrected', 
         freqdevscale=1.5, timedevscale=1.5, action='apply', display='both', flagbackup=False)
</source>
  
Although RFlag has done a pretty good job of finding the bad data, some still remains.  One way to excise the remaining bad data is to use the mode='extend' feature in flagdata, which can extend flags along a chosen axis and removes 'islands' of small data patches in the midst of flagged data.  First, we will extend the flags across polarization, so if any one polarization is flagged, all data for that time / channel will be flagged:

<source lang="python">
# In CASA
flagdata(vis='SNR_G55_10s-hanning.ms', mode='extend', spw='0', extendpols=True,
         action='apply', display='', flagbackup=False)
</source>

Now, we will extend the flags in time and frequency, using the "growtime" and "growfreq" parameters.  For the data here, the <tt>rflag</tt> algorithm seems most likely to miss RFI which should be flagged along more of the time axis, so we will try with growtime=50.0, which will flag all data for a given channel if more than 50% of that channel's time is already flagged, and growfreq=90.0, which will flag the entire spectrum for an integration if more than 90% of the channels in that integration are already flagged.  

<source lang="python">
# In CASA
flagdata(vis='SNR_G55_10s-hanning.ms', mode='extend', spw='0', growtime=50.0, 
         growfreq=90.0, action='apply', display='', flagbackup=False)
</source>

''' spw 1, 2, and 3 '''

Now, let's work on SPW's 1,2, and 3. We've chosen display='both', in order to first review the flags, and decide if too much or too litle is being flagged. We can always click on 'Quit', and decide to change the freqdevscale and timedevscale parameters, as we did with spectral window 0. Note that spw 2 has these parameters lower than spw 1/3, as it contains more RFI. To allow the flagging to continue, click on "Stop Display".

<source lang="python">
# In CASA
flagdata(vis='SNR_G55_10s-hanning.ms', mode='rflag',  spw='1,3', datacolumn='corrected', 
         freqdevscale=4.0, timedevscale=4.0, action='apply', display='both', flagbackup=False)

flagdata(vis='SNR_G55_10s-hanning.ms', mode='extend', spw='1,3', extendpols=True,
         action='apply', display='', flagbackup=False)

flagdata(vis='SNR_G55_10s-hanning.ms', mode='extend', spw='1,3', growtime=50.0, 
         growfreq=90.0, action='apply', display='', flagbackup=False)


flagdata(vis='SNR_G55_10s-hanning.ms', mode='rflag',  spw='2', datacolumn='corrected', 
         freqdevscale=2.5, timedevscale=2.5, action='apply', display='both', flagbackup=False)

flagdata(vis='SNR_G55_10s-hanning.ms', mode='extend', spw='2', extendpols=True,
         action='apply', display='', flagbackup=False)

flagdata(vis='SNR_G55_10s-hanning.ms', mode='extend', spw='2', growtime=50.0, 
         growfreq=90.0, action='apply', display='', flagbackup=False)
</source>

Let's create an after RFlag plot to see the improvements.

[[Image:amp_v_freq.before_after_RFlag.gif|250px|thumb|right|Before and after RFlag for scan 190. (click on plot to see the comparison)]]

<source lang="python">
# In CASA
plotms(vis='SNR_G55_10s-hanning.ms', scan='190' , antenna='ea24',
xaxis='freq', yaxis='amp', ydatacolumn='corrected', coloraxis='spw', title='After RFlag',
correlation='RR,LL', plotfile='amp_v_freq.afterRFlag.png', plotrange=[1.2,2,0,2.5])
</source>

We will now plot amplitude vs. baseline, and look for outliers which we may be able to flag. 

<source lang="python">
# In CASA
plotms(vis='SNR_G55_10s-hanning.ms', scan='30,75,120,165,190,235,303', xaxis='baseline', yaxis='amp', 
ydatacolumn='corrected', iteraxis='scan', correlation='RR,LL', coloraxis = 'baseline')
</source>

It appears that there are a few baselines which have consistently higher-amplitude than the others, indicating that it's probably contaminated by RFI.  

Use the plotms tools to identify this baseline (see above how to do that), which turns out to be ea04&ea08 and ea04&ea21, and is mostly present for spw 1. Let's flag these baselines. 

<source lang="python">
# In CASA
flagdata(vis='SNR_G55_10s-hanning.ms', antenna='ea04&ea08', spw='1', flagbackup=False)
flagdata(vis='SNR_G55_10s-hanning.ms', antenna='ea04&ea21', spw='1', flagbackup=False)
</source>

We will now run flagdata in summary mode to inspect how much data we have flagged thus far. We can also create a plot of the percentage of flagged data with respect to the frequency. This command will also plot the antennas, with circle size representing the percentage of flagged data per antenna. 

<source lang="python">
# In CASA
flagInfo = flagdata(vis='SNR_G55_10s-hanning.ms', mode='summary', action='calculate', display='report', spwchan=T)

print("\n %2.1f%% of G55.7+3.4, %2.1f%% of 3C147, and %2.1f%% of J1925+2106 are flagged. \n" % 
     (100.0 * flagInfo['field']['G55.7+3.4']['flagged'] / flagInfo['field']['G55.7+3.4']['total'], 
     100.0 * flagInfo['field']['0542+498=3C147']['flagged'] / flagInfo['field']['0542+498=3C147']['total'], 
     100.0 * flagInfo['field']['J1925+2106']['flagged'] / flagInfo['field']['J1925+2106']['total']))

print("Spectral windows are flagged as follows:")

for spw in range(0,4):
     print("SPW %s: %2.1f%%" % (spw, 100.0 * flagInfo['spw'][str(spw)]['flagged'] / flagInfo['spw'][str(spw)]['total']))

</source>

So, as a result of the flagging, we have sacrificed almost 35% of the data for G55.7+3.4, and as expected, spectral windows 0 and 2 have the most flagged data.

== Interactive Flagging ==

After plotting our data, it's obvious that there is still some amount of RFI present. We can use plotms to interactively flag points which look bad. A tutorial on flagging data interactively through plotms can be found [https://casaguides.nrao.edu/index.php?title=Data_flagging_with_plotms here]. Please note that flagging data through plotms will not create a backup, so it's important to use the flagmanager before deciding to mark your regions for flagging purposes. 

[[Image:corr.amp_v_freq.interactive_flagging.png|250px|thumb|right|Highlighting and interactively flagging RFI for scan 75.]]

<source lang="python">
# In CASA
flagmanager(vis='SNR_G55_10s.ms', mode='save', 
            versionname='before_interactive_flagging')

plotms(vis='SNR_G55_10s.ms', scan='30,75,120,165,190,235,303', xaxis='frequency', yaxis='amp', 
ydatacolumn='corrected', iteraxis='scan', correlation='RR,LL', coloraxis = 'spw')
</source>

Let's iterate through scans and view scan 75. We can see that spectral window 1 has a large amplitude spike. We can use the "mark regions" button (square with green plus sign) to highlight a majority of spike, and the "locate" button (magnifying glass) to give us information on the highlighted region. We will now click on the "flag" button (red flag) and we should see the plot update, having removed the highlighted region. 

[[Main Page | &#8629; '''CASAguides''']]

-- original: Miriam Hutman <br />
--modifications: Lorant Sjouwerman (4.4.0, 2015/07/07) <br />
--modifications: Jose Salcido (4.5.2, 2016/02/24) <br />
{{Checked 4.5.2}}