Running GUISDAP 8

Conditions

  • You have EISCAT data files (.mat) that you should analyse. In this example we analyse tau2pl_r_cp1_1.10_CP@kir.

Starting guisdap

  • Log on to the analysis machine.
  • You might check that the data directory exists
    
    $ ls /data1/tau2pl_r_cp1_1.10_CP@kir
    20031029_09  20031030_08  20031031_07  20031101_06
    20031029_10  20031030_09  20031031_08  20031101_07
    20031029_11  20031030_10  20031031_09  20031101_08
    20031029_12  20031030_11  20031031_10  20031101_09
    20031029_13  20031030_12  20031031_11  20031101_10
    20031029_14  20031030_13  20031031_12  20031101_11
    20031029_15  20031030_14  20031031_13  20031101_12
    20031029_16  20031030_15  20031031_14  20031101_13
    20031029_17  20031030_16  20031031_15  20031101_14
    20031029_18  20031030_17  20031031_16  20031101_15
    20031029_19  20031030_18  20031031_17  20031101_16
    20031029_20  20031030_19  20031031_18  20031101_17
    20031029_21  20031030_20  20031031_19  20031101_18
    20031029_22  20031030_21  20031031_20  20031101_19
    20031029_23  20031030_22  20031031_21  20031101_20
    20031030_00  20031030_23  20031031_22  20031101_21
    20031030_01  20031031_00  20031031_23  20031101_22
    20031030_02  20031031_01  20031101_00  20031101_23
    20031030_03  20031031_02  20031101_01  20031102_00
    20031030_04  20031031_03  20031101_02  tau2pl_r_cp1_1.10_CP@kir_information
    20031030_05  20031031_04  20031101_03
    20031030_06  20031031_05  20031101_04
    20031030_07  20031031_06  20031101_05
    
    
  • Start up guisdap
    
    $ guisdap -a
    
                                  < M A T L A B >
                      Copyright 1984-2003 The MathWorks, Inc.
                  Version 6.5.1.199709 Release 13 (Service Pack 1)
                                    Aug  4 2003
    
      Using Toolbox Path Cache.  Type "help toolbox_path_cache" for more info.
    
      To get started, type one of these: helpwin, helpdesk, or demo.
      For product information, visit www.mathworks.com.
    
    GUISDAP vs. 8.3 by EISCAT, Lehtinen&Huuskonen;
    GUISDAP vs. 8.3 by EISCAT, Lehtinen&Huuskonen;
    
    
  • A window named “GUISDAP for dummies” should pop up

    gupdum

Fill in the parameters

The window contains the parameters that were used in the last analysis. You might need to change some of them.
First hit the “Reset” button. Guisdap will search for the latest data and set up default parameters. If this is not what you want you have to go further:

  • Data path: Where the raw data files is. Click on the button:

    dpath1   dpath2

    To go up in the hierarchy you need to double-click on the “..” file. Select the right directory and press “OK”. Guisdap will parse the directory name and try to set up everything: Experiment, start/stop time, integration time…
  • Dsp expr: The “radar controller” experiment which is in this example tau2pl (already set up by choosing Data path here). Click on the button and select the right experiment:
  • Site: Select correct site, from which the data comes.
  • Start time: The start time for analysis, year month day hour minute second. Should normally be before the time when the experiment started. A good idea is to set the minutes and seconds to zero. Here:
    
    2003 10 29 9 0 0
    
    
  • Stop time: The stop time for analysis, year month day hour minute second. Should normally be after the time when the experiment stops. A good idea is to set the minutes and seconds to zero. Here:
    
    2003 11 2 24 0 0
    
    
  • Result path: The directory where the results should go. If the last part of it is “AUTO”, it is rewritten using the date, dsp expr, integration time and site. Here:
    
    /analysis/results/2003-10-29_tau2pl_60@kir
    
    

    Normally this is what you want, otherwise click on it and select/create a specific one.

  • Real time: If running the analysis in realtime, this button should be IN. This is also done automatically if the stop time is in the future.
  • Integration time: Number of seconds to integrate, before attempting analysis. For stationary experiments enter a nice number. Here: 120 (=2 minutes). If it is a scanning experiment, you probably want to integrate for as long as the antenna stands still, enter 0.
  • Disp figures: Guisdap produces some plots during the analysis: the data dump, a power profile, the actual fit(s), height profiles, vizu (the parameters vs time). The display is turned on with “1” and off with “0”.
    For remote sites I recommend: “0 0 1 0 1” (fits+vizu)
    For transmitting sites I recommend: “0 1 0 0 1” (powerprofile+vizu)
  • Special: Here you can enter almost any matlab command you want to give before starting analysis. There are hundreds of parameters, but here follows a list of the most common ones
    “analysis_txlimit=0; f_ch_Pt=1.2e6” will integrate the data even if the transmitter is recorded as off and force the tx power to 1.2 MW. Needed sometimes at remote sites.
    “a_Offsetppd” should be set to the number of microseconds the remote site clock differs from Tromso.”f_calTemp” will set the calibration temperature to the correct value, if recorded wrong.”Magic_const” is another parameter to set here, if needed.”analysis_code” choses which part of the data you want to analyse. Needed sometimes for dual-beam or sliced experiments.

    gupdum2

    Remember that “%” will comment out the rest of the line, so one can keep the command there for later reference if you don’t want it this time.
  • When ready, press GO

Analysis

  • The output might look something like this:
    
    29-Oct-2003 09:10:55-09:11:00 integrated
    canon: /opt/guisdap8/exps/tau2pl/tau2plRinit
    canon: /opt/guisdap8/matfiles/pldfvv
    Intersection: [Width Rise]=[52 48] us Veff=7.967e+08
    Warning: 3 points may not be enough for reliable variance determination
    
    ** The model uses the IRI-2001 model at the tx position**
     alt   Ne/1e11     Ti      Te/Ti    coll/1e3    vel   [O+] resid status
    292.5 2.19:0.21 1614:176 1.97:0.31  0.0: 0.0   39: 59 0.95  0.49  OK
    Offset: NaN us  canon: /analysis/results/2003-10-29_tau2pl_60@kir/26039459
    
    29-Oct-2003 09:11:00-09:12:00 integrated
     alt   Ne/1e11     Ti      Te/Ti    coll/1e3    vel   [O+] resid status
    292.5 1.37:0.05 2356:114 0.98:0.07  0.0: 0.0 -163: 25 0.95  1.31  OK
    Offset: 0 us  canon: /analysis/results/2003-10-29_tau2pl_60@kir/26039519
    
    29-Oct-2003 09:12:00-09:13:00 integrated
     alt   Ne/1e11     Ti      Te/Ti    coll/1e3    vel   [O+] resid status
    292.5 2.89:0.05 2099: 51 1.42:0.05  0.0: 0.0  -41: 13 0.95  1.48  OK
    Offset: 0 us  canon: /analysis/results/2003-10-29_tau2pl_60@kir/26039579
    
    29-Oct-2003 09:13:00-09:14:00 integrated
     alt   Ne/1e11     Ti      Te/Ti    coll/1e3    vel   [O+] resid status
    292.5 2.81:0.06 2134: 52 1.45:0.05  0.0: 0.0  -41: 13 0.95  1.35  OK
    Offset: 0 us  canon: /analysis/results/2003-10-29_tau2pl_60@kir/26039639
    
    29-Oct-2003 09:14:00-09:15:00 integrated
     alt   Ne/1e11     Ti      Te/Ti    coll/1e3    vel   [O+] resid status
    292.5 3.60:0.07 1872: 39 1.70:0.05  0.0: 0.0  -19: 10 0.95  1.19  OK
    Offset: 0 us  canon: /analysis/results/2003-10-29_tau2pl_60@kir/26039699

    fits

    Here also the plot of the fit was chosen. Check that the green line (theory) is in phase with the data. A measured offset is displayed, but should be zero. Otherwise, you need to change the “a_Offsetppd” parameter: restart the analysis and do it.

  • If you are running in realtime, the analysis will pause when reaching the end of the data. Here the vizu window will show up, and the program gets each new dump as it arrives. The timeout for waiting for new data is presently set to 5 minutes. If you don’t want to wait when reaching the end of the experiment, you can relase the RT button, and the analysis will finish.
  • When the analysis is finished the vizu plot is saved together with the data.
    Created /analysis/results/2003-11-01_tau2pl_60@kir/2003-11-01_tau2pl_60@kir.eps and .png

    vizu

  • When you are finished, hit the Quit button.

Problems

  • GUSDAP might accidentaly stop, due to unforseeable errors. Sometimes, such interrupted analysis can be continued with the command “anacont”.
  • Report any problems or shortcomings of this manual to Ingemar.