Cookbook

Using LinuxPops to Analyse Spectral Line Data (Quick Start Guide)

Starting:

• Open a terminal and set it's width to: 132 columns, and ~30 rows.
• Log into modelo/smtoast as 'obs' & provide your initials.
• Type: startlinuxpops
• You will see the start-up info displayed. The important info is the listing of defined environments:
  • 0 = Continuum;
  • 1 = Filters; or Forbes-Filters data at SMT
  • 2 = MAC; or AOS data at SMT
  • 3 = AROWS; 12 meter only
  • ...

Note:

• The interactive parser uses minimum matching to distinguish commands. But, variables must be spelled out fully.

Selection Criteria:

• Type: switch 1, to switch to and process filterbank data.
• Type: params, anytime to see what the search criteria are. (shortcut: control-p)
• You can modify the search parameters in various ways:
  • freq 210.153 (without any args, a GUI will pop up to select from list)
  • source Oriona (without any args, a GUI will pop up to select from list)
  • bscan 12 or bscan * to reset to all
  • escan 66 or escan * to reset to all
  • date 2016.0323 or date * to reset to all
  • Select spectro backend using one of the following:
    • feed 1 to select only scans ending in .01; or
    • bkres 1.0 in MHz; or
    • backend F1M-HU or F1M-*U Wildcards allowed
  • many more...
• Type: list to list all scans that meets criteria (shortcut: control-l).
• Type: list all or list /a to list all scans on file if you don't see any with plain list.
• Type: qreject to show a table of rejected scans and why. Useful if list produces no results.

Displaying Scans:

• To fetch a single scan type: get 12.03
• To show a fetched scan type: show
• Type: pshow 12.01 will show scans 12.01 & 12.02 in parallel mode; (or 12.03 for 12M data)
  • Once you have a parallel scan displayed you can do a cross-correlation on that data by typing: xcorrelate
• Type: qshow 12.01 will show scans 12.01, 12.02, 12.03 & 12.04 in quad mode;

Quick Scan Review:

• Type: review to step through and display one scan at a time; hit the space bar to move on to the next scan.
  • NOTE: Hitting 'y' while reviewing scans will add it to the ignored'scan list, then display the next.
  • Type 'q' at any time to quit reviewing scans.
• preview and qreview will review the plots in parallel and quad plots respectfully.
• Type: greview to pop up a GUI of all matching scans; (Spectral Line Only) Click on a mini-plot to see it full screen on the main plot window. Press the Ignore button to ignore any selected bad scans.

Accumulating Data:

You can accumulate all data in the 'stack' several ways.

• In feed mode:
  • c1, c2, c3, c4, cb, will display those IF's.
  • psc12, psc13, psc24, psc34; Stands for Parallel Scans Combine + IF's 1&2, 1&3, etc. This will display a parallel display of the stack.
  • Note: There is a halves command to add them together; or use cb;
• In backend name or bkres matching mode:
  • Use accum to combine all matching scans.
• NOTE:
  • When accumulating scans, the first scan sets the parameters for all subsequent scans that follow:
    • i.e. RA, Dec, Offsets, number of channels, velocity etc.
    • Once set, these dictate whether all subsequent scans meet these additional criteria for summing.
  • So, make sure the first scan matches your needs.
    • Set bscan=1234 to force scan 1234 to be the first accum scan
    • Ignore any scans, such as test scans prior to science scans.
      • ignore 123.01 will tell LinuxPops to ignore scan 123.01 (must include fractional part)
      • The ignore command accepts multiple arguments.
  • Set ignoremappos=1 to ignore small map offsets; (Use with Caution)
  • Set ignorepos=1 to ignore subtle changes to RA, Dec and Vel; (Think Planets, Comets, etc)

Baselines:

• Type bset and enter the number of baselines segments you need. (Limit 64)
  • Click the mouse on the left and then the right side of any baseline regions you wish to define.
    • Two clicks are required for each baseline segment you requested.
    • Note: You are allowed to select beyond the left or right side of the plot to select the first or last channel.
• Type: baseline to remove a simple linear baseline offset
• For more complex baselines:
  • Set nfit to the 'order' of the polynomial elements
    • example: nfit=4
  • Type: gracefit to compute and plot the potential baseline fit.
  • If you are satisfied with the fitting:
    • Type: rmbase to remove the computed baseline from the plot data.
  • If not, then adjust the nfit value and try again. (nfit must be 0 - 19)

Datafiles and Version Control:

You can set the input data file like this:

• filename sdd_file
  • If you don't give a argument, then a popup GUI lets you browse for one.
• NOTE: Environments 0, 1 & 2 are intended for online files.
• gfilename gsdd_file to set the gains file.
  • Again, no args produces a browser GUI.
  • type: gset to automatically set the gains file to the matching data-file.

You can change file versions with two commands:

• fver 98 changes the input file to version 098; ( with no arg, a pop-up GUI will appear to select a version)
• versions pops up a GUI to select a version
• gset sets the gains file to match the datafile; (they must be both in the same directory)

Plot Appearance:

You can modify the way the data plots look:

• Plot Grid Lines:
  • grid /x Turns on vertical X axis lines
  • grid /y Turns on horizontal Y axis lines
  • grid /b Turns on both vertical & horizontal axis lines
  • grid /o Turns off all axis lines
• Plot Axis:
  • fv Sets X axis to Freq-Offset top and Velocity Offset bottom.
  • vf Sets X axis to Velocity Offset top and Freq-Offset bottom.
  • fc Sets X axis to Freq-Offset top and Channel Number bottom.
  • ff, cc, cf, vc & cv also available.
• Baseline Region Markers:
  • bmark=1 boxes appear about zero axis
  • bmark=-1 places boxes at bottom of plot;
  • bmark=0 turns them off.
• Zero Line:
  • zline=1 will draw a line through 0.0K on plot

Performing FFTs:

• Type: fft to perform a fft on the current data array
• Type: fft pick to remove multiple individual points along the fft.
• Type: fft cut to remove a range of points.

Aliases:

There are aliases defined to allow short cuts in commands.

Currently defined aliases are:

Type: alias /l for a listing of available aliases.