Eirik Hatlen / ctools

Commit 000585e98407602de6136f2dfbfac96538b0a8d0

Authored by Luigi Tibaldo
1 parent 3223bbb1

improved description of SUBDIVSQRT and SIGNIFICANCE algortithms for csresmap and csrespec

Inline Side-by-side
Showing 2 changed files with 15 additions and 4 deletions
doc/source/users/reference_manual/csresmap.rst
  View file @000585e
... ... @@ -21,8 +21,11 @@ energies. Four options exist then for residual computation:
21 21 - the subtraction of the model from the counts (SUB)
22 22 - the subtraction and division by the model (SUBDIV)
23 23 - the subtraction and division by the square root of the model (SUBDIVSQRT)
24   -- the computation of the significance using Li & Ma (SIGNIFICANCE)
  24 +- the computation of the significance based on the likelihood-ratio
  25 + test for Poisson statistics (SIGNIFICANCE)
25 26  
  27 +The SUBDIVSQRT and SIGNIFICANCE algorithms approximate the residual
  28 +significance and may become inaccurate in the low-counting regime.
26 29 The residual map is written into a FITS file.
27 30  
28 31  
... ... @@ -108,7 +111,9 @@ General parameters
108 111  
109 112 ``SUBDIVSQRT``: :math:`(DATA - MODEL) / \sqrt{MODEL}`
110 113  
111   - ``SIGNIFICANCE``: :math:`{\rm sign}(DATA-MODEL) \times \sqrt{ 2 \times ( DATA \times \ln \left(\frac{DATA}{MODEL} \right) + MODEL - DATA ) }`
  114 + ``SIGNIFICANCE``: :math:`{\rm sign}(DATA-MODEL) \times \sqrt{ 2
  115 + \times ( DATA \times \ln \left(\frac{DATA}{MODEL} \right) +
  116 + MODEL - DATA ) }`
112 117  
113 118  
114 119 Standard parameters
... ...
doc/source/users/reference_manual/csresspec.rst
  View file @000585e
... ... @@ -35,7 +35,11 @@ Four options exist then for residual computation:
35 35 - the subtraction of the model from the counts (SUB)
36 36 - the subtraction and division by the model (SUBDIV)
37 37 - the subtraction and division by the square root of the model (SUBDIVSQRT)
38   -- the computation of the significance using Li & Ma (SIGNIFICANCE)
  38 +- the computation of the significance based on the likelihood-ratio
  39 + test for Poisson statistics (SIGNIFICANCE)
  40 +
  41 +The SUBDIVSQRT and SIGNIFICANCE algorithms approximate the residual
  42 +significance and may become inaccurate in the low-counting regime.
39 43  
40 44 The counts, model, and residual spectra are written into a FITS file.
41 45  
... ... @@ -162,7 +166,9 @@ General parameters
162 166  
163 167 ``SUBDIVSQRT``: :math:`(DATA - MODEL) / \sqrt{MODEL}`
164 168  
165   - ``SIGNIFICANCE``: :math:`{\rm sign}(DATA-MODEL) \times \sqrt{ 2 \times ( DATA \times \ln \left(\frac{DATA}{MODEL} \right) + MODEL - DATA ) }`
  169 + ``SIGNIFICANCE``: :math:`{\rm sign}(DATA-MODEL) \times \sqrt{ 2
  170 + \times ( DATA \times \ln \left(\frac{DATA}{MODEL} \right) +
  171 + MODEL - DATA ) }`
166 172  
167 173  
168 174 Standard parameters
... ...