.. _hess_source_search:

************************
H.E.S.S. Source Searches
************************

Gamma-ray point sources observed by `H.E.S.S. <https://www.mpi-hd.mpg.de/hfm/HESS/>`_ at TeV energies are candidate targets for this analysis.  The scripts used to test for correlation with these sources are within the ``pev-photons/TeVCat`` directory.

-----------------------
Individual Source Tests
-----------------------

To fit each H.E.S.S. source location indiviually, run:

.. code-block:: bash

    python individual_ts.py

This produces the file ``{prefix}/TeVCat/hess_sources_fit_results.npy``, which stores the TS, :math:`n_s`, and :math:`\gamma` for each source location.  Convert these test statistics to pre-trial p-values with:

.. code-block:: bash

    python hess_ts_to_pvalue.py

Note that this requires the unique declination background trials produce in the ``all_sky`` section.  Add ``--use_original_trials`` to use the trials we generated instead.  Finally, to calculate the sensitivity to each H.E.S.S. source at 1 PeV, given an unbroken power law with the best fit spectral index from H.E.S.S. data, run:

.. code-block:: bash

    python hess_sensitivities.py --ncpu *desired_ncpu*

Combining the information from this section should give you results consistent with the following table:

.. list-table:: Individual H.E.S.S. Source Fit Results
   :widths: auto
   :header-rows: 1

   * - Source
     - :math:`\delta` [:math:`^{\circ}`]
     - :math:`\alpha` [:math:`^{\circ}`]
     - pre-trial p-value 
     - :math:`n_s` 
     - :math:`\gamma`
     - Sens. (x :math:`10^{-20}`)
   * - HESS J1356-645
     - -64.50
     - 209.00
     - ...
     - 0.0
     - ...
     - 3.72
   * - HESS J1507-622
     - -62.35
     - 226.72
     - 0.29
     - 2.5
     - 1.3
     - 4.75
   * - SNR G292.2-00.5
     - -61.40
     - 169.75
     - 0.30
     - 10.0
     - 4.0
     - 8.97
   * - Kookaburra (Rabbit)
     - -60.98
     - 214.52
     - 0.31
     - 3.7
     - 1.8
     - 4.96
   * - HESS J1458-608
     - -60.87
     - 224.54
     - 0.22
     - 9.1
     - 2.4
     - 1.88
   * - HESS J1427-608
     - -60.85
     - 216.97
     - 0.07
     - 25.9
     - 3.2
     - 12.10
   * - Kookaburra (PWN)
     - -60.76
     - 215.04
     - 0.49
     - 1.8
     - 2.2
     - 4.31
   * - SNR G318.2+00.1
     - -59.47
     - 224.44
     - ...
     - 0.0
     - ...
     - 7.79
   * - MSH 15-52
     - -59.16
     - 228.53
     - 0.52
     - 1.6
     - 3.8
     - 14.97
   * - HESS J1018-589 B
     - -58.93
     - 154.74
     - 0.07
     - 23.7
     - 4.0
     - 4.75
   * - HESS J1018-589 A
     - -58.98
     - 154.13
     - 0.08
     - 22.1
     - 4.0
     - 5.22
   * - HESS J1503-582
     - -58.23
     - 225.91
     - 0.17
     - 21.9
     - 4.0
     - 12.73
   * - HESS J1026-582
     - -58.20
     - 156.66
     - 0.09
     - 26.7
     - 4.0
     - 2.50
   * - HESS J1026-582
     - -58.20
     - 156.66
     - 0.09
     - 26.7
     - 4.0
     - 2.50
   * - Westerlund 2
     - -57.79
     - 155.85
     - 0.08
     - 32.0
     - 4.0
     - 9.35
   * - SNR G327.1-01.1
     - -55.08
     - 238.65
     - ...
     - 0.0
     - ...
     - 14.30

-------------
Stacking Test
-------------

Run the stacked likelihood test for all H.E.S.S. source combined by executing:

.. code-block:: bash

    python stacking_test.py

The results are stored in ``{prefix}/TeVCat/stacking_fit_result.npy``. To generate background trials use:

.. code-block:: bash

    python stacking_test.py --bg_trials 100000 --ncpu *desired_ncpu*

The results should match the following table:

.. list-table:: H.E.S.S. Catalog Stacking Result
   :widths: auto
   :header-rows: 1

   * - Test Statistic
     - p-value 
     - :math:`n_s` 
     - :math:`\gamma`
   * - 1.74
     - 0.08
     - 65.23
     - 3.66

---------------------------------
Skymap including H.E.S.S. Sources
---------------------------------

To produce a plot of the all-sky scan p-value map with H.E.S.S. source positions overlaid run:

.. code-block:: bash

    python plot_HESS_region.py

.. figure:: _static/HESS_srcs_w_labels.png
   :scale: 50 %
   :alt: H.E.S.S. Source Skymap

   **Figure 10**:  All-sky likelihood scan pre-trial p-value shown in galactic coordinates. The H.E.S.S. sources in the analysis field of view are shown in black.

-----------------------
Individual Source Limit
-----------------------

Plot best fit spectrum of H.E.S.S. J1356-645 compared to the IceCube upper limit with:

.. code-block:: bash

    python plot_hess_sed.py

.. figure:: _static/hess_source_5.png
   :scale: 50 %
   :alt: H.E.S.S. J1356-645

   **Figure 7 (right)**: Measurements of the source H.E.S.S. J1356-645.  The best fit power law spectrum for the H.E.S.S. data is shown in blue.  The shaded region denotes statistical uncertainty, while the systematic uncertainty is represented by the dashed lines.  The 90\% upper limit set by this analysis is shown in red.