CTAClassicalAnalysis.py

Martin Pierrick, 05/02/2013 01:57 PM

Download (3 KB)

 
1 
def ClassicalAnalysis():
2 
    """
3 
    Typical script for classical analysis of TeV data
4 
    (virtual at this stage, only to help organizing the implementation)
5
6 
    Parameters
7 
    - ....
8 
    - ....
9
10 
    Notes
11 
    - ...
12 
    - ...
13 
    """
14
15 
    # Define source
16 
    # (some parameters to be used only for simulation case)
17 
    src_name=
18 
    src_ra=
19 
    src_dec=
20 
    src_flux=
21 
    src_index=
22 
    src_emin=
23 
    src_emax=
24 
    src_data=''
25
26 
    # Simulate or load data
27 
    if (os.path.exists(src_data)):
28 
            obs = GObservations('obs.xml')
29 
    else:
30 
            ...
31
32 
    # Set analysis parameters
33 
    # (positions passed as GSkyDir objects)
34 
    on_centre=
35 
    on_radius=
36 
    off_centre=
37 
    off_radius=
38 
    off_type='Reflected' or 'Mirror' or 'Ring'
39 
    off_pars=
40 
    acc_curve=
41
42 
    # Set ON region
43 
    # (limited to circular regions at that stage)
44 
    on  = GRegions()
45 
    on.append(GRegionCircle(on_centre,on_radius))
46
47 
    # Set OFF regions
48 
    # (do we want to have the OFF regions catalog build internally in the GOnOffFitter object ?)
49 
    # (each OFF region must be tied to a given pointing so an attribute like pointing ID should be set in each region)
50 
    # A single OFF region
51 
    off.append(GRegionCircle(off_centre,off_radius))
52 
    # Multiple OFF regions defined from the pointings in the observations container, typically for reflected method
53 
    for run in obs:
54 
            off_centre=
55 
            off_radius=on_radius
56 
            off.append(GRegionCircle(off_centre,off_radius))
57
58 
    # Define energy binning
59 
    ebds = GEbounds(10, GEnergy(src_emin, "TeV"), GEnergy(src_emax, "TeV"))
60
61 
    # Initialize analysis object
62 
    onoff= GOnOffFitter(obs,on,off,ebds)
63
64 
    # Make spectra for ON and OFF regions
65 
    # (will create a node function internally, in counts and not intensity, need to check that this is no problem)
66 
    # (here, can have the alternatives of summing spectra over all regions or not, either through different methods or through a flag)
67 
    onoff.make_spectrum()
68 
    onoff.make_spectra()
69
70 
    # Compute all other relevant quantities for ON and OFF regions
71 
    # (how do we handle acceptance right now ? through GCTAModelRadialAcceptance extracted from obs object ?)
72 
    onoff.compute_exposure()
73
74 
    # Can be done from GCTAModelRadialAcceptance.radial().eval(offset)
75 
    # Can start with just the value at the ventre of the OFF region, without integrating over the area
76 
    onoff.compute_acceptance()
77
78 
    # Compute source flux points and significativity
79 
    # (will create a node function internally)
80 
    opt = GOptimizerLM()
81 
    onoff.optimize(opt)
82
83 
    # Alternative: fit some spectral function instead of making a node spectrum
84 
    # (check if/how a spectral model can be handled alone, without a spatial model)
85 
    # (also, do we want to define two models, one for ON and one for OFF ?)
86 
    model=GModelSpectrumPlaw(...)
87 
    onoff2= GOnOffFitter(obs,on,off,ebds,model)
88 
    onoff2.make_spectrum()
89 
    onoff2.compute_exposure()
90 
    onoff2.compute_acceptance()
91 
    opt = GOptimizerLM()
92 
    onoff2.optimize(opt)
93