Commit f5880b0de8348ec8a81275ea9a2f8016a8f1cd2f
1 parent
696c32c2
Update documentation on CTA response handling
Showing
2 changed files
with
146 additions
and
4 deletions
doc/source/user_manual/getting_started/irfs.jpg
0 → 100644
27.6 KB
doc/source/user_manual/getting_started/response.rst
| ... | ... | @@ -18,7 +18,31 @@ CTA Instrument Response Functions |
| 18 | 18 | What are instrument response functions? |
| 19 | 19 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 20 | 20 | |
| 21 | -(will provide a general description of what are IRFs) | |
| 21 | +The instrument response functions provide a mathematical description that | |
| 22 | +links the measured quantities :math:`\vec{d}` of an event to the physical | |
| 23 | +quantities :math:`\vec{p}` of the incident photon. The following figure | |
| 24 | +illustrates this relationship: | |
| 25 | + | |
| 26 | +.. figure:: irfs.jpg | |
| 27 | + :width: 70% | |
| 28 | + | |
| 29 | +:math:`I(\vec{p})` is the gamma-ray intensity arriving at Earth as a | |
| 30 | +function of photon properties :math:`\vec{p}` | |
| 31 | +(which usually are true photon energy, true photon incident direction, | |
| 32 | +and true photon arrival time), | |
| 33 | +while :math:`e(\vec{d})` is the expected event rate as function of event | |
| 34 | +properties :math:`\vec{d}` (which usually are the measured photon energy, | |
| 35 | +measured or reconstructued photon incident direction, and measured photon | |
| 36 | +arrival time). The expected event rate is obtained by integrating the | |
| 37 | +product of | |
| 38 | +the instrumental response function :math:`R(\vec{d}|\vec{p},\vec{a}`) | |
| 39 | +and the emitted intensity :math:`I(\vec{p})` over the photon properties | |
| 40 | +:math:`\vec{p}`. | |
| 41 | +The argument :math:`\vec{a}` in the response function comprises any | |
| 42 | +auxiliary parameter on which the response function may depend on (e.g. | |
| 43 | +pointing direction, triggered telescopes, optical efficiencies, | |
| 44 | +atmospheric conditions, etc.). All these quantities and hence the | |
| 45 | +instrument response function may depend on time. | |
| 22 | 46 | |
| 23 | 47 | |
| 24 | 48 | Data formats |
| ... | ... | @@ -26,8 +50,126 @@ Data formats |
| 26 | 50 | |
| 27 | 51 | So far the data format of the CTA instrument response functions has not |
| 28 | 52 | yet been defined, and information about instrument response functions is |
| 29 | -provided in various ways. | |
| 53 | +provided in various ways. Three types of data formats are so far supported: | |
| 54 | + | |
| 55 | +- :ref:`sec_cta_rsp_perftable` | |
| 56 | + | |
| 57 | +- :ref:`sec_cta_rsp_xspec` | |
| 58 | + | |
| 59 | +- :ref:`sec_cta_rsp_rsptable` | |
| 60 | + | |
| 61 | +All formats provide the instrument response as function of true (and | |
| 62 | +sometimes also measured) photon energy, typically from about 20 GeV to | |
| 63 | +about 125 TeV. | |
| 64 | +Performance tables provide the instrument response for on-axis sources. | |
| 65 | +ARF, RMF and PSF files provide the instrument response for a specific | |
| 66 | +source position (and region) within the field of view. | |
| 67 | +Response tables provide the instrument response as function of position in | |
| 68 | +the field of view. Response tables are thus the most universal form of | |
| 69 | +instrument response functions available, and we recommend using this form | |
| 70 | +for any analysis. | |
| 71 | + | |
| 72 | +.. note :: | |
| 73 | + | |
| 74 | + Instrument response functions are so far only available for a fixed | |
| 75 | + zenith angle of 20 deg. Therefore, no zenith or azimuth angle dependence has | |
| 76 | + been implented in the CTA response functions, but the dependence can be | |
| 77 | + handled by cutting the data into short time segments (typically of 30 | |
| 78 | + minutes in length), and by specifying specific response functions for | |
| 79 | + each segment. | |
| 80 | + | |
| 81 | + | |
| 82 | +Installing the CTA calibration database | |
| 83 | +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | |
| 84 | + | |
| 85 | +After `downloading | |
| 86 | +<https://portal.cta-observatory.org/WG/DM/DM_wiki/DATA_Access/Pages/Science%20Tools.aspx>`_ | |
| 87 | +the latest calibration database (only possible for CTA consortium members), | |
| 88 | +the database is installed using | |
| 89 | + | |
| 90 | +.. code-block:: bash | |
| 91 | + | |
| 92 | + $ [sudo] tar -C $CALDB -zxvf cta-caldb-20140216.tar.gz | |
| 93 | + | |
| 94 | +or | |
| 95 | + | |
| 96 | +.. code-block:: bash | |
| 97 | + | |
| 98 | + $ [sudo] tar -C $CALDB -xvf cta-caldb-20140216.tar | |
| 99 | + | |
| 100 | +depending on whether the database has been retrieved as a gzipped file or | |
| 101 | +not. Depending on your access rights to the ``$CALDB`` directory, | |
| 102 | +installation of the calibration database may require root privileges | |
| 103 | +(type ``sudo`` in this case, otherwise omit this part of the command). | |
| 104 | + | |
| 105 | +The calibration database must be installed into the directory to which you | |
| 106 | +``CALDB`` environment variable points. By default, ctools sets this | |
| 107 | +enviroment variable to | |
| 108 | + | |
| 109 | +.. code-block:: bash | |
| 110 | + | |
| 111 | + $CTOOLS/share/caldb | |
| 112 | + | |
| 113 | +but it may be that some other software installation or configuration | |
| 114 | +setting on your system overwrites this location. If in doubt, type | |
| 115 | + | |
| 116 | +.. code-block:: bash | |
| 30 | 117 | |
| 118 | + echo $CALDB | |
| 119 | + | |
| 120 | +to find out to which directory your ``CALDB`` environment variable is set. | |
| 121 | +You may always overwrite this setting using | |
| 122 | + | |
| 123 | +.. code-block:: bash | |
| 124 | + | |
| 125 | + export CALDB=/my/preferred/caldb/directory | |
| 126 | + | |
| 127 | +which you can add to your ``.bashrc`` file to set the directory | |
| 128 | +permanently. | |
| 129 | +ctools will use the ``CALDB`` environment variable to find out where the | |
| 130 | +calibration database is located. If needed, this mechanism can be | |
| 131 | +circumvented (see :ref:`sec_cta_rsp_abspath`). | |
| 132 | + | |
| 133 | + | |
| 134 | +Specifying the CTA Instrument Response Functions | |
| 135 | +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | |
| 136 | + | |
| 137 | +The specification of the CTA Instrument Response Functions depends on the | |
| 138 | +way how ctools are used. Common to all methods is that the IRFs are | |
| 139 | +defined by a response name and a calibration database name. The latter | |
| 140 | +may in some cases be the path to a directory on your filesystem. | |
| 141 | + | |
| 142 | +There are different means to specify the CTA Instrument Response Functions | |
| 143 | +when using ctools, and the following section describe the | |
| 144 | + | |
| 145 | +Using individual event files | |
| 146 | +^^^^^^^^^^^^^^^^^^^^^^^^^^^^ | |
| 147 | + | |
| 148 | +(to be written) | |
| 149 | + | |
| 150 | + | |
| 151 | +Using observation definition files | |
| 152 | +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ | |
| 153 | + | |
| 154 | +(to be written) | |
| 155 | + | |
| 156 | + | |
| 157 | +From within a Python script | |
| 158 | +^^^^^^^^^^^^^^^^^^^^^^^^^^^ | |
| 159 | + | |
| 160 | +(to be written) | |
| 161 | + | |
| 162 | + | |
| 163 | +.. _sec_cta_rsp_abspath: | |
| 164 | + | |
| 165 | +Using absolute path names to instrument response files | |
| 166 | +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | |
| 167 | + | |
| 168 | +(to be written) | |
| 169 | + | |
| 170 | + | |
| 171 | +Data format details | |
| 172 | +~~~~~~~~~~~~~~~~~~~ | |
| 31 | 173 | |
| 32 | 174 | .. _sec_cta_rsp_perftable: |
| 33 | 175 | |
| ... | ... | @@ -78,8 +220,8 @@ Below an example of a CTA performance table:: |
| 78 | 220 | |
| 79 | 221 | .. _sec_cta_rsp_xspec: |
| 80 | 222 | |
| 81 | -1DC Data Format | |
| 82 | -^^^^^^^^^^^^^^^ | |
| 223 | +ARF, RMF and PSF files | |
| 224 | +^^^^^^^^^^^^^^^^^^^^^^ | |
| 83 | 225 | |
| 84 | 226 | Instrument response information for the first CTA Data Challenge (1DC) has been |
| 85 | 227 | provided in a format that was heavily inspired by the | ... | ... |