ObservationTable

class gammapy.data.ObservationTable(data=None, masked=None, names=None, dtype=None, meta=None, copy=True, rows=None, copy_indices=True, **kwargs)

Bases: astropy.table.Table

Observation table.

This is an Table sub-class, with a few convenience methods. The format of the observation table is described in Observation lists.

Attributes Summary

pointing_galactic	Pointing positions as Galactic (SkyCoord)
pointing_radec	Pointing positions as ICRS (SkyCoord)

Methods Summary

get_obs_idx(obs_id)	Get row index for given obs_id.
read(filename, **kwargs)	Read an observation table from file.
select_linspace_subset(num)	Select subset of observations.
select_obs_id(obs_id)	Get ObservationTable containing only obs_id.
select_observations([selection])	Select subset of observations.
select_range(selection_variable, value_range)	Make an observation table, applying some selection.
select_time_range(selection_variable, time_range)	Make an observation table, applying a time selection.
summary([file])	Info string (str)

Attributes Documentation

pointing_galactic

Pointing positions as Galactic (SkyCoord)

pointing_radec

Pointing positions as ICRS (SkyCoord)

Methods Documentation

get_obs_idx(obs_id)

Get row index for given obs_id.

Raises KeyError if observation is not available.

Parameters:

obs_id : int, list

observation ids

Returns:

idx : list

indices corresponding to obs_id

classmethod read(filename, **kwargs)

Read an observation table from file.

Parameters:

filename : Path, str

Filename

select_linspace_subset(num)

Select subset of observations.

This is mostly useful for testing, if you want to make the analysis run faster.

Parameters:

num : int

Number of samples to select.

Returns:

table : ObservationTable

Subset observation table (a copy).

select_obs_id(obs_id)

Get ObservationTable containing only obs_id.

Raises KeyError if observation is not available.

Parameters:

obs_id: int, list

observation ids

select_observations(selection=None)

Select subset of observations.

Returns a new observation table representing the subset.

There are 3 main kinds of selection criteria, according to the value of the type keyword in the selection dictionary:

• sky regions (boxes or circles)
• time intervals (min, max)
• intervals (min, max) on any other parameter present in the observation table, that can be casted into an Quantity object

Allowed selection criteria are interpreted using the following keywords in the selection dictionary under the type key.

• sky_box and sky_circle are 2D selection criteria acting on sky coordinates

  • sky_box is a squared region delimited by the lon and lat keywords: both tuples of format (min, max); uses select_sky_box
  • sky_circle is a circular region centered in the coordinate marked by the lon and lat keywords, and radius radius; uses select_sky_circle

  in each case, the coordinate system can be specified by the frame keyword (built-in Astropy coordinate frames are supported, e.g. icrs or galactic); an aditional border can be defined using the border keyword

• time_box is a 1D selection criterion acting on the observation start time (TSTART); the interval is set via the time_range keyword; uses select_time_range

• par_box is a 1D selection criterion acting on any parameter defined in the observation table that can be casted into an Quantity object; the parameter name and interval can be specified using the keywords variable and value_range respectively; min = max selects exact values of the parameter; uses select_range

In all cases, the selection can be inverted by activating the inverted flag, in which case, the selection is applied to keep all elements outside the selected range.

A few examples of selection criteria are given below.

Parameters:

selection : dict

Dictionary with a few keywords for applying selection cuts.

Returns:

obs_table : ObservationTable

Observation table after selection.

Examples

>>> selection = dict(type='sky_box', frame='icrs',
...                  lon=Angle([150, 300], 'deg'),
...                  lat=Angle([-50, 0], 'deg'),
...                  border=Angle(2, 'deg'))
>>> selected_obs_table = obs_table.select_observations(selection)

>>> selection = dict(type='sky_circle', frame='galactic',
...                  lon=Angle(0, 'deg'),
...                  lat=Angle(0, 'deg'),
...                  radius=Angle(5, 'deg'),
...                  border=Angle(2, 'deg'))
>>> selected_obs_table = obs_table.select_observations(selection)

>>> selection = dict(type='time_box',
...                  time_range=Time(['2012-01-01T01:00:00', '2012-01-01T02:00:00']))
>>> selected_obs_table = obs_table.select_observations(selection)

>>> selection = dict(type='par_box', variable='ALT',
...                  value_range=Angle([60., 70.], 'deg'))
>>> selected_obs_table = obs_table.select_observations(selection)

>>> selection = dict(type='par_box', variable='OBS_ID',
...                  value_range=[2, 5])
>>> selected_obs_table = obs_table.select_observations(selection)

>>> selection = dict(type='par_box', variable='N_TELS',
...                  value_range=[4, 4])
>>> selected_obs_table = obs_table.select_observations(selection)

select_range(selection_variable, value_range, inverted=False)

Make an observation table, applying some selection.

Generic function to apply a 1D box selection (min, max) to a table on any variable that is in the observation table and can be casted into a Quantity object.

If the range length is 0 (min = max), the selection is applied to the exact value indicated by the min value. This is useful for selection of exact values, for instance in discrete variables like the number of telescopes.

If the inverted flag is activated, the selection is applied to keep all elements outside the selected range.

Parameters:

selection_variable : str

Name of variable to apply a cut (it should exist on the table).

value_range : Quantity-like

Allowed range of values (min, max). The type should be consistent with the selection_variable.

inverted : bool, optional

Invert selection: keep all entries outside the (min, max) range.

Returns:

obs_table : ObservationTable

Observation table after selection.

select_time_range(selection_variable, time_range, inverted=False)

Make an observation table, applying a time selection.

Apply a 1D box selection (min, max) to a table on any time variable that is in the observation table. It supports both fomats: absolute times in Time variables and [MET].

If the inverted flag is activated, the selection is applied to keep all elements outside the selected range.

Parameters:

selection_variable : str

Name of variable to apply a cut (it should exist on the table).

time_range : Time

Allowed time range (min, max).

inverted : bool, optional

Invert selection: keep all entries outside the (min, max) range.

Returns:

obs_table : ObservationTable

Observation table after selection.

summary(file=None)

Info string (str)